public inbox for [email protected]  
help / color / mirror / Atom feed
[PATCH v24 3/5] Make archiver process an auxiliary process
281+ messages / 6 participants
[nested] [flat]

* [PATCH v24 3/5] Make archiver process an auxiliary process
@ 2018-11-07 07:53 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2018-11-07 07:53 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.
Archiver process must be a auxiliary process since it uses shared
memory after stats data wes moved onto shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/bootstrap/bootstrap.c   |  8 +++
 src/backend/postmaster/pgarch.c     | 98 +++++++----------------------
 src/backend/postmaster/pgstat.c     |  6 ++
 src/backend/postmaster/postmaster.c | 35 ++++++++---
 src/include/miscadmin.h             |  2 +
 src/include/pgstat.h                |  1 +
 src/include/postmaster/pgarch.h     |  4 +-
 7 files changed, 67 insertions(+), 87 deletions(-)

diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 8ea033610d..6e38f9a3d2 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -329,6 +329,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			case BgWriterProcess:
 				statmsg = pgstat_get_backend_desc(B_BG_WRITER);
 				break;
+			case ArchiverProcess:
+				statmsg = pgstat_get_backend_desc(B_ARCHIVER);
+				break;
 			case CheckpointerProcess:
 				statmsg = pgstat_get_backend_desc(B_CHECKPOINTER);
 				break;
@@ -456,6 +459,11 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			BackgroundWriterMain();
 			proc_exit(1);		/* should never return */
 
+		case ArchiverProcess:
+			/* don't set signals, archiver has its own agenda */
+			PgArchiverMain();
+			proc_exit(1);		/* should never return */
+
 		case CheckpointerProcess:
 			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index f84f882c4c..4342ebdab4 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -77,7 +77,6 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
@@ -96,7 +95,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
 static void pgarch_exit(SIGNAL_ARGS);
 static void ArchSigHupHandler(SIGNAL_ARGS);
 static void ArchSigTermHandler(SIGNAL_ARGS);
@@ -114,75 +112,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -222,8 +151,8 @@ pgarch_forkexec(void)
  *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
  *	since we don't use 'em, it hardly matters...
  */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -255,8 +184,27 @@ PgArchiverMain(int argc, char *argv[])
 static void
 pgarch_exit(SIGNAL_ARGS)
 {
-	/* SIGQUIT means curl up and die ... */
-	exit(1);
+	PG_SETMASK(&BlockSig);
+
+	/*
+	 * We DO NOT want to run proc_exit() callbacks -- we're here because
+	 * shared memory may be corrupted, so we don't want to try to clean up our
+	 * transaction.  Just nail the windows shut and get out of town.  Now that
+	 * there's an atexit callback to prevent third-party code from breaking
+	 * things by calling exit() directly, we have to reset the callbacks
+	 * explicitly to make this work as intended.
+	 */
+	on_exit_reset();
+
+	/*
+	 * Note we do exit(2) not exit(0).  This is to force the postmaster into a
+	 * system reset cycle if some idiot DBA sends a manual SIGQUIT to a random
+	 * backend.  This is necessary precisely because we don't clean up our
+	 * shared memory state.  (The "dead man switch" mechanism in pmsignal.c
+	 * should ensure the postmaster sees this as a crash, too, but no harm in
+	 * being doubly sure.)
+	 */
+	exit(2);
 }
 
 /* SIGHUP signal handler for archiver process */
diff --git a/src/backend/postmaster/pgstat.c b/src/backend/postmaster/pgstat.c
index fabcf31de8..8299d2a435 100644
--- a/src/backend/postmaster/pgstat.c
+++ b/src/backend/postmaster/pgstat.c
@@ -2932,6 +2932,9 @@ pgstat_bestart(void)
 			case StartupProcess:
 				lbeentry.st_backendType = B_STARTUP;
 				break;
+			case ArchiverProcess:
+				beentry->st_backendType = B_ARCHIVER;
+				break;
 			case BgWriterProcess:
 				lbeentry.st_backendType = B_BG_WRITER;
 				break;
@@ -4275,6 +4278,9 @@ pgstat_get_backend_desc(BackendType backendType)
 
 	switch (backendType)
 	{
+		case B_ARCHIVER:
+			backendDesc = "archiver";
+			break;
 		case B_AUTOVAC_LAUNCHER:
 			backendDesc = "autovacuum launcher";
 			break;
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 9ff2832c00..84fda38249 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -146,7 +146,8 @@
 #define BACKEND_TYPE_AUTOVAC	0x0002	/* autovacuum worker process */
 #define BACKEND_TYPE_WALSND		0x0004	/* walsender process */
 #define BACKEND_TYPE_BGWORKER	0x0008	/* bgworker process */
-#define BACKEND_TYPE_ALL		0x000F	/* OR of all the above */
+#define BACKEND_TYPE_ARCHIVER	0x0010	/* archiver process */
+#define BACKEND_TYPE_ALL		0x001F	/* OR of all the above */
 
 #define BACKEND_TYPE_WORKER		(BACKEND_TYPE_AUTOVAC | BACKEND_TYPE_BGWORKER)
 
@@ -539,6 +540,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1785,7 +1787,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3042,7 +3044,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3187,10 +3189,8 @@ reaper(SIGNAL_ARGS)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3438,7 +3438,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3643,6 +3643,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3915,6 +3927,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5190,7 +5203,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5475,6 +5488,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index bc6e03fbc7..1f4db67f3f 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -399,6 +399,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -411,6 +412,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/pgstat.h b/src/include/pgstat.h
index fe076d823d..65713abc2b 100644
--- a/src/include/pgstat.h
+++ b/src/include/pgstat.h
@@ -718,6 +718,7 @@ typedef struct PgStat_GlobalStats
  */
 typedef enum BackendType
 {
+	B_ARCHIVER,
 	B_AUTOVAC_LAUNCHER,
 	B_AUTOVAC_WORKER,
 	B_BACKEND,
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index 2474eac26a..88f16863d4 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
-- 
2.23.0


----Next_Part(Tue_Dec__3_17_27_59_2019_887)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v24-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH 3/5] Make archiver process an auxiliary process
@ 2018-11-07 07:53 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2018-11-07 07:53 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.
Archiver process must be a auxiliary process since it uses shared
memory after stats data wes moved onto shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/bootstrap/bootstrap.c   |  8 +++
 src/backend/postmaster/pgarch.c     | 98 +++++++----------------------
 src/backend/postmaster/pgstat.c     |  6 ++
 src/backend/postmaster/postmaster.c | 43 ++++++++-----
 src/include/miscadmin.h             |  2 +
 src/include/pgstat.h                |  1 +
 src/include/postmaster/pgarch.h     |  4 +-
 7 files changed, 70 insertions(+), 92 deletions(-)

diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index bfc629c753..2f38420313 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -326,6 +326,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			case StartupProcess:
 				statmsg = pgstat_get_backend_desc(B_STARTUP);
 				break;
+			case ArchiverProcess:
+				statmsg = pgstat_get_backend_desc(B_ARCHIVER);
+				break;
 			case BgWriterProcess:
 				statmsg = pgstat_get_backend_desc(B_BG_WRITER);
 				break;
@@ -451,6 +454,11 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			StartupProcessMain();
 			proc_exit(1);		/* should never return */
 
+		case ArchiverProcess:
+			/* don't set signals, archiver has its own agenda */
+			PgArchiverMain();
+			proc_exit(1);		/* should never return */
+
 		case BgWriterProcess:
 			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index 3ca30badb2..6441e69e9a 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -78,7 +78,6 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
@@ -95,7 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
 static void pgarch_exit(SIGNAL_ARGS);
 static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
@@ -111,75 +109,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -219,8 +148,8 @@ pgarch_forkexec(void)
  *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
  *	since we don't use 'em, it hardly matters...
  */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -252,8 +181,27 @@ PgArchiverMain(int argc, char *argv[])
 static void
 pgarch_exit(SIGNAL_ARGS)
 {
-	/* SIGQUIT means curl up and die ... */
-	exit(1);
+	PG_SETMASK(&BlockSig);
+
+	/*
+	 * We DO NOT want to run proc_exit() callbacks -- we're here because
+	 * shared memory may be corrupted, so we don't want to try to clean up our
+	 * transaction.  Just nail the windows shut and get out of town.  Now that
+	 * there's an atexit callback to prevent third-party code from breaking
+	 * things by calling exit() directly, we have to reset the callbacks
+	 * explicitly to make this work as intended.
+	 */
+	on_exit_reset();
+
+	/*
+	 * Note we do exit(2) not exit(0).  This is to force the postmaster into a
+	 * system reset cycle if some idiot DBA sends a manual SIGQUIT to a random
+	 * process.  This is necessary precisely because we don't clean up our
+	 * shared memory state.  (The "dead man switch" mechanism in pmsignal.c
+	 * should ensure the postmaster sees this as a crash, too, but no harm in
+	 * being doubly sure.)
+	 */
+	exit(2);
 }
 
 /* SIGUSR1 signal handler for archiver process */
diff --git a/src/backend/postmaster/pgstat.c b/src/backend/postmaster/pgstat.c
index 51c486bebd..ca5c6376e5 100644
--- a/src/backend/postmaster/pgstat.c
+++ b/src/backend/postmaster/pgstat.c
@@ -2927,6 +2927,9 @@ pgstat_bestart(void)
 			case StartupProcess:
 				lbeentry.st_backendType = B_STARTUP;
 				break;
+			case ArchiverProcess:
+				beentry->st_backendType = B_ARCHIVER;
+				break;
 			case BgWriterProcess:
 				lbeentry.st_backendType = B_BG_WRITER;
 				break;
@@ -4273,6 +4276,9 @@ pgstat_get_backend_desc(BackendType backendType)
 
 	switch (backendType)
 	{
+		case B_ARCHIVER:
+			backendDesc = "archiver";
+			break;
 		case B_AUTOVAC_LAUNCHER:
 			backendDesc = "autovacuum launcher";
 			break;
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7a92dac525..ee41d7009e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -146,7 +146,8 @@
 #define BACKEND_TYPE_AUTOVAC	0x0002	/* autovacuum worker process */
 #define BACKEND_TYPE_WALSND		0x0004	/* walsender process */
 #define BACKEND_TYPE_BGWORKER	0x0008	/* bgworker process */
-#define BACKEND_TYPE_ALL		0x000F	/* OR of all the above */
+#define BACKEND_TYPE_ARCHIVER	0x0010	/* archiver process */
+#define BACKEND_TYPE_ALL		0x001F	/* OR of all the above */
 
 #define BACKEND_TYPE_WORKER		(BACKEND_TYPE_AUTOVAC | BACKEND_TYPE_BGWORKER)
 
@@ -539,6 +540,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1785,7 +1787,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3042,7 +3044,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3177,20 +3179,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3438,7 +3436,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3643,6 +3641,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3915,6 +3925,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5190,7 +5201,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5475,6 +5486,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 62d64aa0a1..dc7dd1c164 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -397,6 +397,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -409,6 +410,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/pgstat.h b/src/include/pgstat.h
index 36b530bc27..0334213b98 100644
--- a/src/include/pgstat.h
+++ b/src/include/pgstat.h
@@ -718,6 +718,7 @@ typedef struct PgStat_GlobalStats
  */
 typedef enum BackendType
 {
+	B_ARCHIVER,
 	B_AUTOVAC_LAUNCHER,
 	B_AUTOVAC_WORKER,
 	B_BACKEND,
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
-- 
2.23.0


----Next_Part(Wed_Jan_22_17_24_04_2020_620)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v31 4/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index d62c12310a..2f8672ac0c 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 5480a024e0..d398ce6f03 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -319,6 +319,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -439,30 +442,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index 37be0e2bbb..063d1323ea 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	pqsignal(SIGQUIT, SignalHandlerForCrashExit);
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 73d278f3b2..e97995f973 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -539,6 +539,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1785,7 +1786,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3055,7 +3056,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3190,20 +3191,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3451,7 +3448,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3656,6 +3653,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3928,6 +3937,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5208,7 +5218,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5251,16 +5261,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5493,6 +5493,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 9938cddb57..af4599bd82 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -187,6 +187,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 9ec7b31cce..4435df82b6 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -313,6 +313,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 14fa127ab1..619b2f9c71 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index d21780108b..a87e7dc711 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -272,6 +272,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.2


----Next_Part(Fri_Apr__3_17_31_17_2020_104)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v31-0005-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v43 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index f39dc4ddf1..d69e20961d 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index a7ed93fdc1..edda899554 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index ed1b65358d..e3a520def9 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 5d09822c81..8a4706a870 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -547,6 +547,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1787,7 +1788,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3041,7 +3042,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3176,20 +3177,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3437,7 +3434,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3642,6 +3639,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3938,6 +3947,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5185,7 +5195,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5230,16 +5240,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5481,6 +5481,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 7dc3911590..fc23539137 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -181,6 +181,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 221af87e71..65443063e8 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 72e3352398..bba002ad24 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -418,6 +418,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -430,6 +431,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index e77f76ae8a..a656910d02 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Dec_11_16_50_03_2020_915)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v43-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v40 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index cae93ab69d..6908bec2f9 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index a7ed93fdc1..edda899554 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index ed1b65358d..e3a520def9 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 959e3b8873..b811c961a6 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -555,6 +555,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1800,7 +1801,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3054,7 +3055,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3189,20 +3190,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3450,7 +3447,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3655,6 +3652,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3951,6 +3960,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5230,7 +5240,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5275,16 +5285,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5526,6 +5526,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 88566bd9fa..746bed773e 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -181,6 +181,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 221af87e71..65443063e8 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 72e3352398..bba002ad24 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -418,6 +418,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -430,6 +431,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 9c9a50ae45..de20520b8c 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -345,6 +345,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.4


----Next_Part(Wed_Nov__4_17_39_10_2020_208)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v40-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v32 4/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index d62c12310a..2f8672ac0c 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 5480a024e0..d398ce6f03 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -319,6 +319,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -439,30 +442,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index 37be0e2bbb..063d1323ea 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	pqsignal(SIGQUIT, SignalHandlerForCrashExit);
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 73d278f3b2..e97995f973 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -539,6 +539,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1785,7 +1786,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3055,7 +3056,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3190,20 +3191,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3451,7 +3448,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3656,6 +3653,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3928,6 +3937,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5208,7 +5218,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5251,16 +5261,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5493,6 +5493,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 9938cddb57..af4599bd82 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -187,6 +187,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 9ec7b31cce..4435df82b6 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -313,6 +313,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 14fa127ab1..619b2f9c71 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index d21780108b..a87e7dc711 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -272,6 +272,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.2


----Next_Part(Tue_Apr__7_16_38_17_2020_299)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v32-0005-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v42 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index cae93ab69d..6908bec2f9 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index a7ed93fdc1..edda899554 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index ed1b65358d..e3a520def9 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index b7799ed1d2..a1dd6964e9 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -547,6 +547,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1785,7 +1786,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3039,7 +3040,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3174,20 +3175,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3435,7 +3432,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3640,6 +3637,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3936,6 +3945,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5177,7 +5187,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5222,16 +5232,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5473,6 +5473,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 88566bd9fa..746bed773e 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -181,6 +181,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 221af87e71..65443063e8 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 72e3352398..bba002ad24 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -418,6 +418,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -430,6 +431,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 9c9a50ae45..de20520b8c 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -345,6 +345,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.4


----Next_Part(Wed_Nov_11_10_07_22_2020_796)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v42-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v25 4/8] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/bootstrap/bootstrap.c   | 22 +++++----
 src/backend/postmaster/pgarch.c     | 75 +----------------------------
 src/backend/postmaster/postmaster.c | 43 +++++++++++------
 src/include/miscadmin.h             |  2 +
 src/include/postmaster/pgarch.h     |  4 +-
 5 files changed, 46 insertions(+), 100 deletions(-)

diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 5480a024e0..d398ce6f03 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -319,6 +319,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -439,30 +442,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index 37be0e2bbb..4971b3ae42 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -78,7 +78,6 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
@@ -95,7 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
 static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
@@ -110,75 +108,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -218,8 +147,8 @@ pgarch_forkexec(void)
  *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
  *	since we don't use 'em, it hardly matters...
  */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 2b9ab32293..cab7fb5381 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -146,7 +146,8 @@
 #define BACKEND_TYPE_AUTOVAC	0x0002	/* autovacuum worker process */
 #define BACKEND_TYPE_WALSND		0x0004	/* walsender process */
 #define BACKEND_TYPE_BGWORKER	0x0008	/* bgworker process */
-#define BACKEND_TYPE_ALL		0x000F	/* OR of all the above */
+#define BACKEND_TYPE_ARCHIVER	0x0010	/* archiver process */
+#define BACKEND_TYPE_ALL		0x001F	/* OR of all the above */
 
 #define BACKEND_TYPE_WORKER		(BACKEND_TYPE_AUTOVAC | BACKEND_TYPE_BGWORKER)
 
@@ -539,6 +540,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1785,7 +1787,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3055,7 +3057,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3190,20 +3192,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3451,7 +3449,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3656,6 +3654,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3928,6 +3938,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5208,7 +5219,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5493,6 +5504,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 14fa127ab1..619b2f9c71 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
-- 
2.18.2


----Next_Part(Thu_Mar_19_20_30_04_2020_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v25-0005-Use-latch-instead-of-SIGUSR1-to-wake-up-archiver.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v38 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index cae93ab69d..6908bec2f9 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 76b2f5066f..81bfaea869 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index ed1b65358d..e3a520def9 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 959e3b8873..b811c961a6 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -555,6 +555,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1800,7 +1801,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3054,7 +3055,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3189,20 +3190,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3450,7 +3447,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3655,6 +3652,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3951,6 +3960,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5230,7 +5240,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5275,16 +5285,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5526,6 +5526,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 88566bd9fa..746bed773e 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -181,6 +181,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 221af87e71..65443063e8 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 72e3352398..bba002ad24 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -418,6 +418,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -430,6 +431,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 9c9a50ae45..de20520b8c 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -345,6 +345,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.4


----Next_Part(Thu_Oct__1_09_07_22_2020_252)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v38-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v34 4/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index cdd586fcfb..ee3444284b 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 5480a024e0..d398ce6f03 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -319,6 +319,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -439,30 +442,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index 37be0e2bbb..063d1323ea 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	pqsignal(SIGQUIT, SignalHandlerForCrashExit);
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 160afe9f39..9de9396628 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -539,6 +539,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1785,7 +1786,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3068,7 +3069,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3203,20 +3204,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3464,7 +3461,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3669,6 +3666,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3941,6 +3950,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5216,7 +5226,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5259,16 +5269,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5501,6 +5501,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index f5eef6fa4e..46f0198510 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -187,6 +187,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index e917dfe92d..0783692c83 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -336,6 +336,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 14fa127ab1..619b2f9c71 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 1ee9000b2b..8ab6d859bb 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -270,6 +270,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.2


----Next_Part(Mon_Jun__1_18_00_01_2020_089)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v34-0005-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v35 4/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index cdd586fcfb..ee3444284b 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 5480a024e0..d398ce6f03 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -319,6 +319,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -439,30 +442,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index 37be0e2bbb..063d1323ea 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	pqsignal(SIGQUIT, SignalHandlerForCrashExit);
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index b4d475bb0b..30f3fc6ba1 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -539,6 +539,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1785,7 +1786,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3068,7 +3069,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3203,20 +3204,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3464,7 +3461,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3669,6 +3666,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3941,6 +3950,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5216,7 +5226,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5259,16 +5269,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5501,6 +5501,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index f5eef6fa4e..46f0198510 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -187,6 +187,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index e917dfe92d..0783692c83 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -336,6 +336,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 18bc8a7b90..cf0b463eba 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 1ee9000b2b..8ab6d859bb 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -270,6 +270,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.2


----Next_Part(Mon_Jun__8_17_32_04_2020_489)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v35-0005-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v37 3/6] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 8f8734dc1d..2e6c322142 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 76b2f5066f..81bfaea869 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index ed1b65358d..e3a520def9 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 959e3b8873..b811c961a6 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -555,6 +555,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1800,7 +1801,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3054,7 +3055,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3189,20 +3190,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3450,7 +3447,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3655,6 +3652,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3951,6 +3960,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5230,7 +5240,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5275,16 +5285,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5526,6 +5526,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 88566bd9fa..746bed773e 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -181,6 +181,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 221af87e71..65443063e8 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 72e3352398..bba002ad24 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -418,6 +418,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -430,6 +431,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 9c9a50ae45..de20520b8c 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -345,6 +345,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.4


----Next_Part(Fri_Sep_25_09_27_26_2020_419)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v37-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v41 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index cae93ab69d..6908bec2f9 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index a7ed93fdc1..edda899554 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index ed1b65358d..e3a520def9 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 959e3b8873..b811c961a6 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -555,6 +555,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1800,7 +1801,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3054,7 +3055,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3189,20 +3190,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3450,7 +3447,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3655,6 +3652,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3951,6 +3960,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5230,7 +5240,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5275,16 +5285,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5526,6 +5526,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 88566bd9fa..746bed773e 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -181,6 +181,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 221af87e71..65443063e8 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 72e3352398..bba002ad24 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -418,6 +418,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -430,6 +431,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 9c9a50ae45..de20520b8c 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -345,6 +345,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.4


----Next_Part(Fri_Nov__6_09_27_56_2020_738)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v41-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v39 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index cae93ab69d..6908bec2f9 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 76b2f5066f..81bfaea869 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index ed1b65358d..e3a520def9 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 959e3b8873..b811c961a6 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -555,6 +555,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1800,7 +1801,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3054,7 +3055,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3189,20 +3190,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3450,7 +3447,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3655,6 +3652,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3951,6 +3960,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5230,7 +5240,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5275,16 +5285,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5526,6 +5526,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 88566bd9fa..746bed773e 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -181,6 +181,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 221af87e71..65443063e8 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 72e3352398..bba002ad24 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -418,6 +418,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -430,6 +431,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 9c9a50ae45..de20520b8c 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -345,6 +345,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.4


----Next_Part(Tue_Oct__6_10_06_44_2020_382)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v39-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v30 4/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlog.c        |  46 +++++++++
 src/backend/access/transam/xlogarchive.c |   2 +-
 src/backend/bootstrap/bootstrap.c        |  22 +++--
 src/backend/postmaster/pgarch.c          | 115 ++++-------------------
 src/backend/postmaster/postmaster.c      |  53 ++++++-----
 src/include/access/xlog.h                |   2 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 10 files changed, 112 insertions(+), 136 deletions(-)

diff --git a/src/backend/access/transam/xlog.c b/src/backend/access/transam/xlog.c
index 977d448f50..d75a257da2 100644
--- a/src/backend/access/transam/xlog.c
+++ b/src/backend/access/transam/xlog.c
@@ -684,6 +684,13 @@ typedef struct XLogCtlData
 	 */
 	Latch		recoveryWakeupLatch;
 
+	/*
+	 * archiverWakeupLatch is used to wake up the archiver process to process
+	 * completed WAL segments, if it is waiting for WAL to arrive.  Protected
+	 * by info_lck.
+	 */
+	Latch	   *archiverWakeupLatch;
+
 	/*
 	 * During recovery, we keep a copy of the latest checkpoint record here.
 	 * lastCheckPointRecPtr points to start of checkpoint record and
@@ -8450,6 +8457,45 @@ GetLastSegSwitchData(XLogRecPtr *lastSwitchLSN)
 	return result;
 }
 
+/*
+ * XLogArchiveWakeupStart - Set up archiver wakeup stuff
+ */
+void
+XLogArchiveWakeupStart(void)
+{
+	SpinLockAcquire(&XLogCtl->info_lck);
+	Assert(XLogCtl->archiverWakeupLatch == NULL);
+	XLogCtl->archiverWakeupLatch = MyLatch;
+	SpinLockRelease(&XLogCtl->info_lck);
+}
+
+/*
+ * XLogArchiveWakeupEnd - Clean up archiver wakeup stuff
+ */
+void
+XLogArchiveWakeupEnd(void)
+{
+	SpinLockAcquire(&XLogCtl->info_lck);
+	XLogCtl->archiverWakeupLatch = NULL;
+	SpinLockRelease(&XLogCtl->info_lck);
+}
+
+/*
+ * XLogWakeupArchiver - Wake up archiver process
+ */
+void
+XLogArchiveWakeup(void)
+{
+	Latch *latch;
+
+	SpinLockAcquire(&XLogCtl->info_lck);
+	latch = XLogCtl->archiverWakeupLatch;
+	SpinLockRelease(&XLogCtl->info_lck);
+
+	if (latch)
+		SetLatch(latch);
+}
+
 /*
  * This must be called ONCE during postmaster or standalone-backend shutdown
  */
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index d62c12310a..d6c8bcce8b 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -490,7 +490,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		XLogArchiveWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 5480a024e0..d398ce6f03 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -319,6 +319,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -439,30 +442,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index 37be0e2bbb..7b66cb5613 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,7 +79,6 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
@@ -95,8 +95,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +108,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +141,16 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Clean up notification stuff on exit */
+static void
+PgArchiverKill(int code, Datum arg)
+{
+	XLogArchiveWakeupEnd();
+}
+
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +162,23 @@ PgArchiverMain(int argc, char *argv[])
 	pqsignal(SIGQUIT, SignalHandlerForCrashExit);
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
-
+	XLogArchiveWakeupStart();
+	on_shmem_exit(PgArchiverKill, 0);
+	
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 73d278f3b2..a4b9d212a2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -146,7 +146,8 @@
 #define BACKEND_TYPE_AUTOVAC	0x0002	/* autovacuum worker process */
 #define BACKEND_TYPE_WALSND		0x0004	/* walsender process */
 #define BACKEND_TYPE_BGWORKER	0x0008	/* bgworker process */
-#define BACKEND_TYPE_ALL		0x000F	/* OR of all the above */
+#define BACKEND_TYPE_ARCHIVER	0x0010	/* archiver process */
+#define BACKEND_TYPE_ALL		0x001F	/* OR of all the above */
 
 #define BACKEND_TYPE_WORKER		(BACKEND_TYPE_AUTOVAC | BACKEND_TYPE_BGWORKER)
 
@@ -539,6 +540,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1785,7 +1787,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3055,7 +3057,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3190,20 +3192,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3451,7 +3449,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3656,6 +3654,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3928,6 +3938,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5208,7 +5219,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5251,16 +5262,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5493,6 +5494,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 9ec7b31cce..c0e149d526 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -313,6 +313,8 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 14fa127ab1..619b2f9c71 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
-- 
2.18.2


----Next_Part(Wed_Apr__1_17_37_23_2020_570)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v30-0005-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v33 4/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index cdd586fcfb..ee3444284b 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 5480a024e0..d398ce6f03 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -319,6 +319,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -439,30 +442,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index 37be0e2bbb..063d1323ea 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	pqsignal(SIGQUIT, SignalHandlerForCrashExit);
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 160afe9f39..9de9396628 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -539,6 +539,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1785,7 +1786,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3068,7 +3069,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3203,20 +3204,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3464,7 +3461,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3669,6 +3666,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3941,6 +3950,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5216,7 +5226,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5259,16 +5269,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5501,6 +5501,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 5aa19d3f78..7acc48734e 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -187,6 +187,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index e917dfe92d..0783692c83 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -336,6 +336,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 14fa127ab1..619b2f9c71 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index ae4f573ab4..1a8a0c2e15 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -272,6 +272,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.2


----Next_Part(Fri_May_15_17_30_36_2020_111)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v33-0005-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v33 4/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 55becd65d4..2417de72a0 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 5480a024e0..d398ce6f03 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -319,6 +319,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -439,30 +442,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index 37be0e2bbb..063d1323ea 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	pqsignal(SIGQUIT, SignalHandlerForCrashExit);
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 0578e92ba9..2687d2af47 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -539,6 +539,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1785,7 +1786,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3055,7 +3056,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3190,20 +3191,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3451,7 +3448,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3656,6 +3653,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3928,6 +3937,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5203,7 +5213,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5246,16 +5256,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5488,6 +5488,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 5aa19d3f78..7acc48734e 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -187,6 +187,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 0a12afb59e..099da8cd76 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -336,6 +336,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 14fa127ab1..619b2f9c71 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index ae4f573ab4..1a8a0c2e15 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -272,6 +272,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.2


----Next_Part(Thu_Apr_30_13_22_05_2020_409)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v33-0005-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v29 4/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlog.c        |  46 ++++++++++
 src/backend/access/transam/xlogarchive.c |   2 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++---
 src/backend/postmaster/pgarch.c          | 102 ++++-------------------
 src/backend/postmaster/postmaster.c      |  53 ++++++------
 src/include/access/xlog.h                |   2 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 10 files changed, 108 insertions(+), 127 deletions(-)

diff --git a/src/backend/access/transam/xlog.c b/src/backend/access/transam/xlog.c
index 977d448f50..d75a257da2 100644
--- a/src/backend/access/transam/xlog.c
+++ b/src/backend/access/transam/xlog.c
@@ -684,6 +684,13 @@ typedef struct XLogCtlData
 	 */
 	Latch		recoveryWakeupLatch;
 
+	/*
+	 * archiverWakeupLatch is used to wake up the archiver process to process
+	 * completed WAL segments, if it is waiting for WAL to arrive.  Protected
+	 * by info_lck.
+	 */
+	Latch	   *archiverWakeupLatch;
+
 	/*
 	 * During recovery, we keep a copy of the latest checkpoint record here.
 	 * lastCheckPointRecPtr points to start of checkpoint record and
@@ -8450,6 +8457,45 @@ GetLastSegSwitchData(XLogRecPtr *lastSwitchLSN)
 	return result;
 }
 
+/*
+ * XLogArchiveWakeupStart - Set up archiver wakeup stuff
+ */
+void
+XLogArchiveWakeupStart(void)
+{
+	SpinLockAcquire(&XLogCtl->info_lck);
+	Assert(XLogCtl->archiverWakeupLatch == NULL);
+	XLogCtl->archiverWakeupLatch = MyLatch;
+	SpinLockRelease(&XLogCtl->info_lck);
+}
+
+/*
+ * XLogArchiveWakeupEnd - Clean up archiver wakeup stuff
+ */
+void
+XLogArchiveWakeupEnd(void)
+{
+	SpinLockAcquire(&XLogCtl->info_lck);
+	XLogCtl->archiverWakeupLatch = NULL;
+	SpinLockRelease(&XLogCtl->info_lck);
+}
+
+/*
+ * XLogWakeupArchiver - Wake up archiver process
+ */
+void
+XLogArchiveWakeup(void)
+{
+	Latch *latch;
+
+	SpinLockAcquire(&XLogCtl->info_lck);
+	latch = XLogCtl->archiverWakeupLatch;
+	SpinLockRelease(&XLogCtl->info_lck);
+
+	if (latch)
+		SetLatch(latch);
+}
+
 /*
  * This must be called ONCE during postmaster or standalone-backend shutdown
  */
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index d62c12310a..d6c8bcce8b 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -490,7 +490,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		XLogArchiveWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 5480a024e0..d398ce6f03 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -319,6 +319,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -439,30 +442,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index 37be0e2bbb..6fe7a136ba 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,7 +79,6 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
@@ -95,8 +95,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +108,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +141,21 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
+/* Clean up notification stuff on exit */
+static void
+PgArchiverKill(int code, Datum arg)
+{
+	XLogArchiveWakeupEnd();
+}
+
 /*
  * PgArchiverMain
  *
  *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
  *	since we don't use 'em, it hardly matters...
  */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,7 +167,7 @@ PgArchiverMain(int argc, char *argv[])
 	pqsignal(SIGQUIT, SignalHandlerForCrashExit);
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
@@ -240,24 +176,14 @@ PgArchiverMain(int argc, char *argv[])
 	MyBackendType = B_ARCHIVER;
 	init_ps_display(NULL);
 
+	XLogArchiveWakeupStart();
+	on_shmem_exit(PgArchiverKill, 0);
+	
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 73d278f3b2..a4b9d212a2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -146,7 +146,8 @@
 #define BACKEND_TYPE_AUTOVAC	0x0002	/* autovacuum worker process */
 #define BACKEND_TYPE_WALSND		0x0004	/* walsender process */
 #define BACKEND_TYPE_BGWORKER	0x0008	/* bgworker process */
-#define BACKEND_TYPE_ALL		0x000F	/* OR of all the above */
+#define BACKEND_TYPE_ARCHIVER	0x0010	/* archiver process */
+#define BACKEND_TYPE_ALL		0x001F	/* OR of all the above */
 
 #define BACKEND_TYPE_WORKER		(BACKEND_TYPE_AUTOVAC | BACKEND_TYPE_BGWORKER)
 
@@ -539,6 +540,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1785,7 +1787,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3055,7 +3057,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3190,20 +3192,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3451,7 +3449,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3656,6 +3654,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3928,6 +3938,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5208,7 +5219,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5251,16 +5262,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5493,6 +5494,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 9ec7b31cce..c0e149d526 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -313,6 +313,8 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 14fa127ab1..619b2f9c71 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
-- 
2.18.2


----Next_Part(Wed_Apr__1_15_15_11_2020_923)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v29-0005-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v44 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index f39dc4ddf1..d69e20961d 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index a7ed93fdc1..edda899554 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index ed1b65358d..e3a520def9 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 5d09822c81..8a4706a870 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -547,6 +547,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1787,7 +1788,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3041,7 +3042,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3176,20 +3177,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3437,7 +3434,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3642,6 +3639,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3938,6 +3947,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5185,7 +5195,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5230,16 +5240,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5481,6 +5481,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 7dc3911590..fc23539137 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -181,6 +181,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 221af87e71..65443063e8 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 72e3352398..bba002ad24 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -418,6 +418,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -430,6 +431,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index e77f76ae8a..a656910d02 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Mon_Dec_21_17_16_20_2020_608)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v44-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v52 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 224 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  83 ++++-----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 9 files changed, 139 insertions(+), 222 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..2558bcfb08 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,13 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +493,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c399e6c267 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,68 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	*latch;		/* latch to wake the archiver up */
+	slock_t  mutex;		/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size	size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool	found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver ", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +144,48 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	*latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +210,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +248,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +270,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +656,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..34c8551288 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_17_51_37_2021_192)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v52-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v36 4/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 8f8734dc1d..2e6c322142 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -489,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 76b2f5066f..81bfaea869 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index 37be0e2bbb..063d1323ea 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	pqsignal(SIGQUIT, SignalHandlerForCrashExit);
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 42223c0f61..4c296dd214 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -554,6 +554,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1799,7 +1800,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3051,7 +3052,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3186,20 +3187,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3447,7 +3444,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3652,6 +3649,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3948,6 +3957,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5223,7 +5233,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5268,16 +5278,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5515,6 +5515,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index aa9fbd8054..9cc5226667 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -181,6 +181,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 221af87e71..65443063e8 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 1c67de2ede..54ce0b97d7 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 72e3352398..bba002ad24 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -418,6 +418,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -430,6 +431,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index b3200874ca..e3ffc63f14 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index 56c5ec4481..c691acf8cd 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 9c9a50ae45..de20520b8c 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -345,6 +345,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.18.4


----Next_Part(Tue_Sep__8_17_55_57_2020_197)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v36-0005-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v47 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 683ab64f76..6cdaf3753d 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -355,6 +355,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_21_12_03_48_2021_284)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v47-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v46 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index c87ffc6549..a1e51c5b99 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Thu_Jan_14_15_14_25_2021_903)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v46-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v45 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 7de27ee4e0..af91c313e2 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1788,7 +1789,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3027,7 +3028,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3162,20 +3163,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3423,7 +3420,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3629,6 +3626,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3932,6 +3941,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5160,7 +5170,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5214,16 +5224,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5465,6 +5465,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index db0cfaa360..aabf9d73eb 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 0786fcf103..430d438303 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -354,6 +354,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Jan__8_10_24_34_2021_185)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v45-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v49 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_16_53_11_2021_575)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v49-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v48 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   6 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++--
 src/backend/postmaster/pgarch.c          | 130 +++--------------------
 src/backend/postmaster/postmaster.c      |  50 +++++----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/access/xlog.h                |   3 +
 src/include/access/xlogarchive.h         |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |   4 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   3 +
 11 files changed, 69 insertions(+), 154 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..d01859bde5 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,9 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
+#include "storage/latch.h"
 #include "storage/pmsignal.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +492,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..9a2e21bf86 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,13 +79,11 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
@@ -95,8 +94,6 @@ static volatile sig_atomic_t ready_to_stop = false;
 static pid_t pgarch_forkexec(void);
 #endif
 
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -110,75 +107,6 @@ static void pgarch_archiveDone(char *xlog);
  * ------------------------------------------------------------
  */
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
 /* ------------------------------------------------------------
  * Local functions called by archiver follow
  * ------------------------------------------------------------
@@ -212,14 +140,9 @@ pgarch_forkexec(void)
 #endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +154,26 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
 	exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
-}
-
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +198,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +236,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +258,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..22e1cd244c 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,16 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 * Was it the archiver?  Normal exit can be ignored; we'll start a new
+		 * one at the next iteration of the postmaster's main loop, if
+		 * necessary. Any other exit condition is treated as a crash.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
 			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			continue;
 		}
 
@@ -3403,7 +3400,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,6 +3606,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) PgArchPID)));
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/*
 	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
 	 * necessary, but it seems like a good idea for robustness, and it
@@ -3912,6 +3921,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5140,7 +5150,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5204,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5445,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/access/xlog.h b/src/include/access/xlog.h
index 75ec1073bd..551f518cc2 100644
--- a/src/include/access/xlog.h
+++ b/src/include/access/xlog.h
@@ -339,6 +339,9 @@ extern XLogRecPtr GetRedoRecPtr(void);
 extern XLogRecPtr GetInsertRecPtr(void);
 extern XLogRecPtr GetFlushRecPtr(void);
 extern XLogRecPtr GetLastImportantRecPtr(void);
+extern void XLogArchiveWakeupStart(void);
+extern void XLogArchiveWakeupEnd(void);
+extern void XLogArchiveWakeup(void);
 extern void RemovePromoteSignalFiles(void);
 
 extern bool PromoteIsTriggered(void);
diff --git a/src/include/access/xlogarchive.h b/src/include/access/xlogarchive.h
index 3edd1a976c..1a59181cf9 100644
--- a/src/include/access/xlogarchive.h
+++ b/src/include/access/xlogarchive.h
@@ -25,6 +25,7 @@ extern void ExecuteRecoveryCommand(const char *command, const char *commandName,
 extern void KeepFileRestoredFromArchive(const char *path, const char *xlogfname);
 extern void XLogArchiveNotify(const char *xlog);
 extern void XLogArchiveNotifySeg(XLogSegNo segno);
+extern void XLogArchiveWakeup(void);
 extern void XLogArchiveForceDone(const char *xlog);
 extern bool XLogArchiveCheckDone(const char *xlog);
 extern bool XLogArchiveIsBusy(const char *xlog);
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..385b002dfe 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -32,8 +32,6 @@
  */
 extern int	pgarch_start(void);
 
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..491747c113 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,9 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
+	/* Current shared estimate of appropriate spins_per_delay value */
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
-- 
2.27.0


----Next_Part(Fri_Mar__5_17_18_56_2021_497)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v48-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v51 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 194 ++++-------------------
 src/backend/postmaster/postmaster.c      |  83 +++++-----
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 94 insertions(+), 236 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..c219780c60 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,13 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,148 +75,29 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +109,27 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +602,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ *  exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	ProcGlobal->archiverLatch = NULL;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..fd509869aa 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3141,24 +3142,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3175,6 +3158,26 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving of remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
+		 * PostmasterStateMachine() will advance to the next shutdown step.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			if (PgArchStartupAllowed())
+				PgArchPID = StartArchiver();
+			continue;
+		}
+
 		/* Was it the system logger?  If so, try to start a new one */
 		if (pid == SysLoggerPID)
 		{
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d71700a565 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -350,6 +350,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* The proc of the Startup process, since not in ProcArray */
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Wed_Mar_10_12_10_39_2021_432)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v51-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v50 3/7] Make archiver process an auxiliary process
@ 2020-03-13 07:59 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2020-03-13 07:59 UTC (permalink / raw)

This is a preliminary patch for shared-memory based stats collector.

Archiver process must be a auxiliary process since it uses shared
memory after stats data was moved into shared-memory. Make the process
an auxiliary process in order to make it work.
---
 src/backend/access/transam/xlogarchive.c |   7 +-
 src/backend/bootstrap/bootstrap.c        |  22 +--
 src/backend/postmaster/pgarch.c          | 174 +++--------------------
 src/backend/postmaster/postmaster.c      |  88 ++++++------
 src/backend/storage/lmgr/proc.c          |   1 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  10 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |  10 +-
 9 files changed, 83 insertions(+), 232 deletions(-)

diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..8b66a30931 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -29,7 +29,8 @@
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
+#include "storage/latch.h"
+#include "storage/proc.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -490,8 +491,8 @@ XLogArchiveNotify(const char *xlog)
 	}
 
 	/* Notify archiver that it's got something to do */
-	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+	if (IsUnderPostmaster && ProcGlobal->archiverLatch)
+		SetLatch(ProcGlobal->archiverLatch);
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..e237cedaff 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -48,6 +48,7 @@
 #include "storage/latch.h"
 #include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -78,25 +79,17 @@
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
@@ -105,121 +98,9 @@ static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
 
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
-
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
-}
-#endif							/* EXEC_BACKEND */
-
-
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,31 +112,24 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->archiverLatch = &MyProc->procLatch;
 
 	pgarch_MainLoop();
 
-	exit(0);
-}
-
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
-{
-	int			save_errno = errno;
-
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	proc_exit(0);
 }
 
 /* SIGUSR2 signal handler for archiver process */
@@ -282,14 +156,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +194,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +216,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..d9f8d82650 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3126,6 +3127,18 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the archiver?  We treat it the same way to WAL receiver.
+		 */
+		if (pid == PgArchPID)
+		{
+			PgArchPID = 0;
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
+			continue;
+		}
+
 		/*
 		 * Was it the autovacuum launcher?	Normal exit can be ignored; we'll
 		 * start a new one at the next iteration of the postmaster's main
@@ -3141,24 +3154,6 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
-		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
-		 * PostmasterStateMachine() will advance to the next shutdown step.
-		 */
-		if (pid == PgArchPID)
-		{
-			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
-			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
-			continue;
-		}
-
 		/*
 		 * Was it the statistics collector?  If so, just try to start a new
 		 * one; no need to force reset of the rest of the system.  (If fail,
@@ -3403,7 +3398,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3604,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3912,6 +3904,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -4989,6 +4982,19 @@ SubPostmasterMain(int argc, char *argv[])
 
 		AuxiliaryProcessMain(argc - 2, argv + 2);	/* does not return */
 	}
+	if (strcmp(argv[1], "--forkarch") == 0)
+	{
+		/* Restore basic shared memory pointers */
+		InitShmemAccess(UsedShmemSegAddr);
+
+		/* Need a PGPROC to run CreateSharedMemoryAndSemaphores */
+		InitAuxiliaryProcess();
+
+		/* Attach process to shared data structures */
+		CreateSharedMemoryAndSemaphores();
+
+		PgArchiverMain(); /* does not return */
+	}
 	if (strcmp(argv[1], "--forkavlauncher") == 0)
 	{
 		/* Restore basic shared memory pointers */
@@ -5037,12 +5043,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5140,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5194,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5435,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 897045ee27..bf5d6da89b 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -182,6 +182,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->archiverLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bdc97e308..adb9f819bb 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -419,6 +419,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -431,6 +432,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..737ed69e57 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,6 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern void PgArchiverMain(void) pg_attribute_noreturn();
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..d73890b228 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	int			startupProcPid;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
 	int			startupBufferPinWaitBufId;
+	/* Archiver process's latch */
+	Latch	   *archiverLatch;
 } PROC_HDR;
 
 extern PGDLLIMPORT PROC_HDR *ProcGlobal;
@@ -370,11 +372,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
-- 
2.27.0


----Next_Part(Tue_Mar__9_18_29_34_2021_806)--
Content-Type: Text/X-Patch; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="v50-0004-Shared-memory-based-stats-collector.patch"



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v55 3/3] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 240 ++++++++++-------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 164 insertions(+), 212 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..07ba989a37 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,101 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int			pgprocno;		/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
-
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
-
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	Size		size = 0;
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
-
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	size = add_size(size, sizeof(PgArchData));
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
+	return size;
+}
 
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-			PgArchiverMain(0, NULL);
-			break;
-#endif
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-		default:
-			return (int) pgArchPid;
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
 	}
-
-	/* shouldn't get here */
-	return 0;
 }
 
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
 /*
- * pgarch_forkexec() -
+ *  PgArchIsSuppressed
  *
- * Format up the arglist for, then fork and exec, archive process
+ * Return true if archiver relaunch is suppressed.
+ *
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-static pid_t
-pgarch_forkexec(void)
+bool
+PgArchCanRestart(void)
 {
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
+	static time_t last_pgarch_start_time = 0;
+	time_t		curtime = time(NULL);
 
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
+	/*
+	 * allow to launch also for the first time and when something's wrong in
+	 * the time values.
+	 */
+	if (last_pgarch_start_time == 0 ||
+		curtime < last_pgarch_start_time ||
+		(curtime - last_pgarch_start_time) >= PGARCH_RESTART_INTERVAL)
+	{
+		last_pgarch_start_time = curtime;
 
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
+		return true;
+	}
 
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +177,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our pgproc index so that backends can use our latch to wake
+	 * us up while we're sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int			arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will
+	 * be relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +246,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +284,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +306,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +692,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..e8af05c04e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||			\
+	  (XLogArchivingAlways() &&									  \
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 PgArchCanRestart())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for
+		 * it to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and
+		 * syslogger processes are disregarded since they are not connected to
+		 * shared memory; we also disregard dead_end children here. Walsenders
+		 * and archiver are also disregarded, they will be terminated later
+		 * after writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..1e47a143e1 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchCanRestart(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_17_24_43_2021_866)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v53 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 227 ++++++++---------------
 src/backend/postmaster/postmaster.c      |  74 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  13 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 137 insertions(+), 218 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..818c9a1f3f 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,153 +71,70 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	Latch	   *latch;			/* latch to wake the archiver up */
+	slock_t		mutex;			/* locks this struct */
+} PgArchData;
+
 
 /* ----------
  * Local data
  * ----------
  */
-static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
 
-/*
- * pgarch_start
- *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
- *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
- */
-int
-pgarch_start(void)
-{
-	time_t		curtime;
-	pid_t		pgArchPid;
+	return size;
+}
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
-	last_pgarch_start_time = curtime;
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
 
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
+	if (!found)
 	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		SpinLockInit(&PgArch->mutex);
+		PgArch->latch = NULL;
 	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
 }
-#endif							/* EXEC_BACKEND */
-
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +146,49 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = MyLatch;
+	SpinLockRelease(&PgArch->mutex);
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	Latch	   *latch;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
+	/* fetching the latch pointer might not be atomic, so use spinlock */
+	SpinLockAcquire(&PgArch->mutex);
+	latch = PgArch->latch;
+	SpinLockRelease(&PgArch->mutex);
 
-	errno = save_errno;
+	if (latch)
+		SetLatch(latch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +213,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +251,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +273,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +659,17 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	SpinLockAcquire(&PgArch->mutex);
+	PgArch->latch = NULL;
+	SpinLockRelease(&PgArch->mutex);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..460e629145 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -548,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1793,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3008,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3143,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3406,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3612,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3804,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3911,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5037,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5134,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5188,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5429,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..d053bf1564 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,9 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_09_23_13_2021_947)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* [PATCH v54 3/7] Make archiver process an auxiliary process
@ 2021-03-11 09:01 Kyotaro Horiguchi <[email protected]>
  0 siblings, 0 replies; 281+ messages in thread

From: Kyotaro Horiguchi @ 2021-03-11 09:01 UTC (permalink / raw)

This makes it possible to archiver process uses ipc functions and let
us monitor archiver process status.
---
 doc/src/sgml/monitoring.sgml             |   1 +
 src/backend/access/transam/xlogarchive.c |   4 +-
 src/backend/bootstrap/bootstrap.c        |  22 ++-
 src/backend/postmaster/pgarch.c          | 241 +++++++++--------------
 src/backend/postmaster/postmaster.c      |  81 ++++----
 src/backend/storage/ipc/ipci.c           |   2 +
 src/include/miscadmin.h                  |   2 +
 src/include/postmaster/pgarch.h          |  14 +-
 src/include/storage/pmsignal.h           |   1 -
 src/include/storage/proc.h               |   8 +-
 src/tools/pgindent/typedefs.list         |   1 +
 11 files changed, 160 insertions(+), 217 deletions(-)

diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 1ba813bbb9..a96455bdb2 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -935,6 +935,7 @@ postgres   27093  0.0  0.0  30096  2752 ?        Ss   11:34   0:00 postgres: ser
        <literal>logical replication worker</literal>,
        <literal>parallel worker</literal>, <literal>background writer</literal>,
        <literal>client backend</literal>, <literal>checkpointer</literal>,
+       <literal>archiver</literal>,
        <literal>startup</literal>, <literal>walreceiver</literal>,
        <literal>walsender</literal> and <literal>walwriter</literal>.
        In addition, background workers registered by extensions may have
diff --git a/src/backend/access/transam/xlogarchive.c b/src/backend/access/transam/xlogarchive.c
index 1c5a4f8b5a..26b023e754 100644
--- a/src/backend/access/transam/xlogarchive.c
+++ b/src/backend/access/transam/xlogarchive.c
@@ -25,11 +25,11 @@
 #include "common/archive.h"
 #include "miscadmin.h"
 #include "postmaster/startup.h"
+#include "postmaster/pgarch.h"
 #include "replication/walsender.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/lwlock.h"
-#include "storage/pmsignal.h"
 
 /*
  * Attempt to retrieve the specified file from off-line archival storage.
@@ -491,7 +491,7 @@ XLogArchiveNotify(const char *xlog)
 
 	/* Notify archiver that it's got something to do */
 	if (IsUnderPostmaster)
-		SendPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER);
+		PgArchWakeup();
 }
 
 /*
diff --git a/src/backend/bootstrap/bootstrap.c b/src/backend/bootstrap/bootstrap.c
index 6f615e6622..41da0c5059 100644
--- a/src/backend/bootstrap/bootstrap.c
+++ b/src/backend/bootstrap/bootstrap.c
@@ -317,6 +317,9 @@ AuxiliaryProcessMain(int argc, char *argv[])
 		case StartupProcess:
 			MyBackendType = B_STARTUP;
 			break;
+		case ArchiverProcess:
+			MyBackendType = B_ARCHIVER;
+			break;
 		case BgWriterProcess:
 			MyBackendType = B_BG_WRITER;
 			break;
@@ -437,30 +440,29 @@ AuxiliaryProcessMain(int argc, char *argv[])
 			proc_exit(1);		/* should never return */
 
 		case StartupProcess:
-			/* don't set signals, startup process has its own agenda */
 			StartupProcessMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
+
+		case ArchiverProcess:
+			PgArchiverMain();
+			proc_exit(1);
 
 		case BgWriterProcess:
-			/* don't set signals, bgwriter has its own agenda */
 			BackgroundWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case CheckpointerProcess:
-			/* don't set signals, checkpointer has its own agenda */
 			CheckpointerMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalWriterProcess:
-			/* don't set signals, walwriter has its own agenda */
 			InitXLOGAccess();
 			WalWriterMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		case WalReceiverProcess:
-			/* don't set signals, walreceiver has its own agenda */
 			WalReceiverMain();
-			proc_exit(1);		/* should never return */
+			proc_exit(1);
 
 		default:
 			elog(PANIC, "unrecognized process type: %d", (int) MyAuxProcType);
diff --git a/src/backend/postmaster/pgarch.c b/src/backend/postmaster/pgarch.c
index edec311f12..8e572afea3 100644
--- a/src/backend/postmaster/pgarch.c
+++ b/src/backend/postmaster/pgarch.c
@@ -38,16 +38,14 @@
 #include "libpq/pqsignal.h"
 #include "miscadmin.h"
 #include "pgstat.h"
-#include "postmaster/fork_process.h"
 #include "postmaster/interrupt.h"
 #include "postmaster/pgarch.h"
-#include "postmaster/postmaster.h"
-#include "storage/dsm.h"
 #include "storage/fd.h"
 #include "storage/ipc.h"
 #include "storage/latch.h"
-#include "storage/pg_shmem.h"
 #include "storage/pmsignal.h"
+#include "storage/procsignal.h"
+#include "storage/spin.h"
 #include "utils/guc.h"
 #include "utils/ps_status.h"
 
@@ -73,6 +71,12 @@
  */
 #define NUM_ORPHAN_CLEANUP_RETRIES 3
 
+/* Shared memory area for archiver process */
+typedef struct PgArchData
+{
+	int		pgprocno;				/* pgprocno of archiver process */
+} PgArchData;
+
 
 /* ----------
  * Local data
@@ -80,146 +84,79 @@
  */
 static time_t last_pgarch_start_time;
 static time_t last_sigterm_time = 0;
+static PgArchData *PgArch = NULL;
 
 /*
  * Flags set by interrupt handlers for later service in the main loop.
  */
-static volatile sig_atomic_t wakened = false;
 static volatile sig_atomic_t ready_to_stop = false;
 
 /* ----------
  * Local function forward declarations
  * ----------
  */
-#ifdef EXEC_BACKEND
-static pid_t pgarch_forkexec(void);
-#endif
-
-NON_EXEC_STATIC void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-static void pgarch_waken(SIGNAL_ARGS);
 static void pgarch_waken_stop(SIGNAL_ARGS);
 static void pgarch_MainLoop(void);
 static void pgarch_ArchiverCopyLoop(void);
 static bool pgarch_archiveXlog(char *xlog);
 static bool pgarch_readyXlog(char *xlog);
 static void pgarch_archiveDone(char *xlog);
+static void pgarch_die(int code, Datum arg);
 
+/* Report shared memory space needed by PgArchShmemInit */
+Size
+PgArchShmemSize(void)
+{
+	Size		size = 0;
 
-/* ------------------------------------------------------------
- * Public functions called from postmaster follow
- * ------------------------------------------------------------
- */
+	size = add_size(size, sizeof(PgArchData));
+
+	return size;
+}
+
+/* Allocate and initialize archiver-related shared memory */
+void
+PgArchShmemInit(void)
+{
+	bool		found;
+
+	PgArch = (PgArchData *)
+		ShmemInitStruct("Archiver Data", PgArchShmemSize(), &found);
+
+	if (!found)
+	{
+		/* First time through, so initialize */
+		MemSet(PgArch, 0, PgArchShmemSize());
+		PgArch->pgprocno = INVALID_PGPROCNO;
+	}
+}
 
 /*
- * pgarch_start
+ *  PgArchIsSuppressed
  *
- *	Called from postmaster at startup or after an existing archiver
- *	died.  Attempt to fire up a fresh archiver process.
+ * Return true if archiver relaunch is suppressed.
  *
- *	Returns PID of child process, or 0 if fail.
- *
- *	Note: if fail, we will be called again from the postmaster main loop.
+ * This is a safety valve to protect against continuous respawn attempts if the
+ * archiver is dying immediately at launch. Note that since we will retry to
+ * launch the archiver from the postmaster main loop, we will get another
+ * chance later.
  */
-int
-pgarch_start(void)
+bool
+PgArchIsSuppressed(void)
 {
-	time_t		curtime;
-	pid_t		pgArchPid;
+	time_t	curtime = time(NULL);
 
-	/*
-	 * Do nothing if no archiver needed
-	 */
-	if (!XLogArchivingActive())
-		return 0;
+	if ((curtime - last_pgarch_start_time) < PGARCH_RESTART_INTERVAL)
+		return true;
 
-	/*
-	 * Do nothing if too soon since last archiver start.  This is a safety
-	 * valve to protect against continuous respawn attempts if the archiver is
-	 * dying immediately at launch. Note that since we will be re-called from
-	 * the postmaster main loop, we will get another chance later.
-	 */
-	curtime = time(NULL);
-	if ((unsigned int) (curtime - last_pgarch_start_time) <
-		(unsigned int) PGARCH_RESTART_INTERVAL)
-		return 0;
 	last_pgarch_start_time = curtime;
-
-#ifdef EXEC_BACKEND
-	switch ((pgArchPid = pgarch_forkexec()))
-#else
-	switch ((pgArchPid = fork_process()))
-#endif
-	{
-		case -1:
-			ereport(LOG,
-					(errmsg("could not fork archiver: %m")));
-			return 0;
-
-#ifndef EXEC_BACKEND
-		case 0:
-			/* in postmaster child ... */
-			InitPostmasterChild();
-
-			/* Close the postmaster's sockets */
-			ClosePostmasterPorts(false);
-
-			/* Drop our connection to postmaster's shared memory, as well */
-			dsm_detach_all();
-			PGSharedMemoryDetach();
-
-			PgArchiverMain(0, NULL);
-			break;
-#endif
-
-		default:
-			return (int) pgArchPid;
-	}
-
-	/* shouldn't get here */
-	return 0;
-}
-
-/* ------------------------------------------------------------
- * Local functions called by archiver follow
- * ------------------------------------------------------------
- */
-
-
-#ifdef EXEC_BACKEND
-
-/*
- * pgarch_forkexec() -
- *
- * Format up the arglist for, then fork and exec, archive process
- */
-static pid_t
-pgarch_forkexec(void)
-{
-	char	   *av[10];
-	int			ac = 0;
-
-	av[ac++] = "postgres";
-
-	av[ac++] = "--forkarch";
-
-	av[ac++] = NULL;			/* filled in by postmaster_forkexec */
-
-	av[ac] = NULL;
-	Assert(ac < lengthof(av));
-
-	return postmaster_forkexec(ac, av);
+	return false;
 }
-#endif							/* EXEC_BACKEND */
 
 
-/*
- * PgArchiverMain
- *
- *	The argc/argv parameters are valid only in EXEC_BACKEND case.  However,
- *	since we don't use 'em, it hardly matters...
- */
-NON_EXEC_STATIC void
-PgArchiverMain(int argc, char *argv[])
+/* Main entry point for archiver process */
+void
+PgArchiverMain(void)
 {
 	/*
 	 * Ignore all signals usually bound to some action in the postmaster,
@@ -231,33 +168,51 @@ PgArchiverMain(int argc, char *argv[])
 	/* SIGQUIT handler was already set up by InitPostmasterChild */
 	pqsignal(SIGALRM, SIG_IGN);
 	pqsignal(SIGPIPE, SIG_IGN);
-	pqsignal(SIGUSR1, pgarch_waken);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
 	pqsignal(SIGUSR2, pgarch_waken_stop);
+
 	/* Reset some signals that are accepted by postmaster but not here */
 	pqsignal(SIGCHLD, SIG_DFL);
+
+	/* Unblock signals (they were blocked when the postmaster forked us) */
 	PG_SETMASK(&UnBlockSig);
 
-	MyBackendType = B_ARCHIVER;
-	init_ps_display(NULL);
+	/* We shouldn't be launched unnecessarily. */
+	Assert(XLogArchivingActive());
+
+	/* Arrange to clean up at archiver exit */
+	on_shmem_exit(pgarch_die, 0);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	PgArch->pgprocno = MyProc->pgprocno;
 
 	pgarch_MainLoop();
 
-	exit(0);
+	proc_exit(0);
 }
 
-/* SIGUSR1 signal handler for archiver process */
-static void
-pgarch_waken(SIGNAL_ARGS)
+/*
+ * Wake up the archiver
+ */
+void
+PgArchWakeup(void)
 {
-	int			save_errno = errno;
+	int	arch_pgprocno = PgArch->pgprocno;
 
-	/* set flag that there is work to be done */
-	wakened = true;
-	SetLatch(MyLatch);
-
-	errno = save_errno;
+	/*
+	 * We don't acquire ProcArrayLock here.  It's actually fine because
+	 * procLatch isn't ever freed, so we just can potentially set the wrong
+	 * process' (or no process') latch.  Even in that case the archiver will be
+	 * relaunched shortly and will start archiving.
+	 */
+	if (arch_pgprocno != INVALID_PGPROCNO)
+		SetLatch(&ProcGlobal->allProcs[arch_pgprocno].procLatch);
 }
 
+
 /* SIGUSR2 signal handler for archiver process */
 static void
 pgarch_waken_stop(SIGNAL_ARGS)
@@ -282,14 +237,6 @@ pgarch_MainLoop(void)
 	pg_time_t	last_copy_time = 0;
 	bool		time_to_stop;
 
-	/*
-	 * We run the copy loop immediately upon entry, in case there are
-	 * unarchived files left over from a previous database run (or maybe the
-	 * archiver died unexpectedly).  After that we wait for a signal or
-	 * timeout before doing more.
-	 */
-	wakened = true;
-
 	/*
 	 * There shouldn't be anything for the archiver to do except to wait for a
 	 * signal ... however, the archiver exists to protect our data, so she
@@ -328,12 +275,8 @@ pgarch_MainLoop(void)
 		}
 
 		/* Do what we're here for */
-		if (wakened || time_to_stop)
-		{
-			wakened = false;
-			pgarch_ArchiverCopyLoop();
-			last_copy_time = time(NULL);
-		}
+		pgarch_ArchiverCopyLoop();
+		last_copy_time = time(NULL);
 
 		/*
 		 * Sleep until a signal is received, or until a poll is forced by
@@ -354,13 +297,9 @@ pgarch_MainLoop(void)
 							   WL_LATCH_SET | WL_TIMEOUT | WL_POSTMASTER_DEATH,
 							   timeout * 1000L,
 							   WAIT_EVENT_ARCHIVER_MAIN);
-				if (rc & WL_TIMEOUT)
-					wakened = true;
 				if (rc & WL_POSTMASTER_DEATH)
 					time_to_stop = true;
 			}
-			else
-				wakened = true;
 		}
 
 		/*
@@ -744,3 +683,15 @@ pgarch_archiveDone(char *xlog)
 	StatusFilePath(rlogdone, xlog, ".done");
 	(void) durable_rename(rlogready, rlogdone, WARNING);
 }
+
+
+/*
+ * pgarch_die
+ *
+ * Exit-time cleanup handler
+ */
+static void
+pgarch_die(int code, Datum arg)
+{
+	PgArch->pgprocno = INVALID_PGPROCNO;
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index edab95a19e..4f8e364284 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -443,9 +443,10 @@ static void InitPostmasterDeathWatchHandle(void);
  * even during recovery.
  */
 #define PgArchStartupAllowed()	\
-	((XLogArchivingActive() && pmState == PM_RUN) ||	\
-	 (XLogArchivingAlways() &&	\
-	  (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY)))
+	(((XLogArchivingActive() && pmState == PM_RUN) ||	\
+	  (XLogArchivingAlways() &&									\
+	   (pmState == PM_RECOVERY || pmState == PM_HOT_STANDBY))) && \
+	 !PgArchIsSuppressed())
 
 #ifdef EXEC_BACKEND
 
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #endif							/* EXEC_BACKEND */
 
 #define StartupDataBase()		StartChildProcess(StartupProcess)
+#define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
@@ -1792,7 +1794,7 @@ ServerLoop(void)
 
 		/* If we have lost the archiver, try to start a new one. */
 		if (PgArchPID == 0 && PgArchStartupAllowed())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/* If we need to signal the autovacuum launcher, do so now */
 		if (avlauncher_needs_signal)
@@ -3007,7 +3009,7 @@ reaper(SIGNAL_ARGS)
 			if (!IsBinaryUpgrade && AutoVacuumingActive() && AutoVacPID == 0)
 				AutoVacPID = StartAutoVacLauncher();
 			if (PgArchStartupAllowed() && PgArchPID == 0)
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			if (PgStatPID == 0)
 				PgStatPID = pgstat_start();
 
@@ -3142,20 +3144,22 @@ reaper(SIGNAL_ARGS)
 		}
 
 		/*
-		 * Was it the archiver?  If so, just try to start a new one; no need
-		 * to force reset of the rest of the system.  (If fail, we'll try
-		 * again in future cycles of the main loop.).  Unless we were waiting
-		 * for it to shut down; don't restart it in that case, and
+		 * Was it the archiver?  If exit status is zero (normal) or one (FATAL
+		 * exit), we assume everything is all right just like normal backends
+		 * and just try to restart a new one so that we immediately retry
+		 * archiving remaining files. (If fail, we'll try again in future
+		 * cycles of the postmaster's main loop.) Unless we were waiting for it
+		 * to shut down; don't restart it in that case, and
 		 * PostmasterStateMachine() will advance to the next shutdown step.
 		 */
 		if (pid == PgArchPID)
 		{
 			PgArchPID = 0;
-			if (!EXIT_STATUS_0(exitstatus))
-				LogChildExit(LOG, _("archiver process"),
-							 pid, exitstatus);
+			if (!EXIT_STATUS_0(exitstatus) && !EXIT_STATUS_1(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("archiver process"));
 			if (PgArchStartupAllowed())
-				PgArchPID = pgarch_start();
+				PgArchPID = StartArchiver();
 			continue;
 		}
 
@@ -3403,7 +3407,7 @@ CleanupBackend(int pid,
 
 /*
  * HandleChildCrash -- cleanup after failed backend, bgwriter, checkpointer,
- * walwriter, autovacuum, or background worker.
+ * walwriter, autovacuum, archiver or background worker.
  *
  * The objectives here are to clean up our local state about the child
  * process, and to signal all other remaining children to quickdie.
@@ -3609,19 +3613,16 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(AutoVacPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
-	/*
-	 * Force a power-cycle of the pgarch process too.  (This isn't absolutely
-	 * necessary, but it seems like a good idea for robustness, and it
-	 * simplifies the state-machine logic in the case where a shutdown request
-	 * arrives during crash processing.)
-	 */
-	if (PgArchPID != 0 && take_action)
+	/* Take care of the archiver too */
+	if (pid == PgArchPID)
+		PgArchPID = 0;
+	else if (PgArchPID != 0 && take_action)
 	{
 		ereport(DEBUG2,
 				(errmsg_internal("sending %s to process %d",
-								 "SIGQUIT",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
 								 (int) PgArchPID)));
-		signal_child(PgArchPID, SIGQUIT);
+		signal_child(PgArchPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
 	/*
@@ -3804,12 +3805,11 @@ PostmasterStateMachine(void)
 		 * (including autovac workers), no bgworkers (including unconnected
 		 * ones), and no walwriter, autovac launcher or bgwriter.  If we are
 		 * doing crash recovery or an immediate shutdown then we expect the
-		 * checkpointer to exit as well, otherwise not. The archiver, stats,
-		 * and syslogger processes are disregarded since they are not
-		 * connected to shared memory; we also disregard dead_end children
-		 * here. Walsenders are also disregarded, they will be terminated
-		 * later after writing the checkpoint record, like the archiver
-		 * process.
+		 * checkpointer to exit as well, otherwise not. The stats and syslogger
+		 * processes are disregarded since they are not connected to shared
+		 * memory; we also disregard dead_end children here. Walsenders and
+		 * archiver are also disregarded, they will be terminated later after
+		 * writing the checkpoint record.
 		 */
 		if (CountChildren(BACKEND_TYPE_ALL - BACKEND_TYPE_WALSND) == 0 &&
 			StartupPID == 0 &&
@@ -3912,6 +3912,7 @@ PostmasterStateMachine(void)
 			Assert(CheckpointerPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
+			Assert(PgArchPID == 0);
 			/* syslogger is not considered here */
 			pmState = PM_NO_CHILDREN;
 		}
@@ -5037,12 +5038,6 @@ SubPostmasterMain(int argc, char *argv[])
 
 		StartBackgroundWorker();
 	}
-	if (strcmp(argv[1], "--forkarch") == 0)
-	{
-		/* Do not want to attach to shared memory */
-
-		PgArchiverMain(argc, argv); /* does not return */
-	}
 	if (strcmp(argv[1], "--forkcol") == 0)
 	{
 		/* Do not want to attach to shared memory */
@@ -5140,7 +5135,7 @@ sigusr1_handler(SIGNAL_ARGS)
 		 */
 		Assert(PgArchPID == 0);
 		if (XLogArchivingAlways())
-			PgArchPID = pgarch_start();
+			PgArchPID = StartArchiver();
 
 		/*
 		 * If we aren't planning to enter hot standby mode later, treat
@@ -5194,16 +5189,6 @@ sigusr1_handler(SIGNAL_ARGS)
 	if (StartWorkerNeeded || HaveCrashedWorker)
 		maybe_start_bgworkers();
 
-	if (CheckPostmasterSignal(PMSIGNAL_WAKEN_ARCHIVER) &&
-		PgArchPID != 0)
-	{
-		/*
-		 * Send SIGUSR1 to archiver process, to wake it up and begin archiving
-		 * next WAL file.
-		 */
-		signal_child(PgArchPID, SIGUSR1);
-	}
-
 	/* Tell syslogger to rotate logfile if requested */
 	if (SysLoggerPID != 0)
 	{
@@ -5445,6 +5430,10 @@ StartChildProcess(AuxProcType type)
 				ereport(LOG,
 						(errmsg("could not fork startup process: %m")));
 				break;
+			case ArchiverProcess:
+				ereport(LOG,
+						(errmsg("could not fork archiver process: %m")));
+				break;
 			case BgWriterProcess:
 				ereport(LOG,
 						(errmsg("could not fork background writer process: %m")));
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index f9bbe97b50..3e4ec53a97 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -144,6 +144,7 @@ CreateSharedMemoryAndSemaphores(void)
 		size = add_size(size, ReplicationOriginShmemSize());
 		size = add_size(size, WalSndShmemSize());
 		size = add_size(size, WalRcvShmemSize());
+		size = add_size(size, PgArchShmemSize());
 		size = add_size(size, ApplyLauncherShmemSize());
 		size = add_size(size, SnapMgrShmemSize());
 		size = add_size(size, BTreeShmemSize());
@@ -258,6 +259,7 @@ CreateSharedMemoryAndSemaphores(void)
 	ReplicationOriginShmemInit();
 	WalSndShmemInit();
 	WalRcvShmemInit();
+	PgArchShmemInit();
 	ApplyLauncherShmemInit();
 
 	/*
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 54693e047a..013850ac28 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -417,6 +417,7 @@ typedef enum
 	BootstrapProcess,
 	StartupProcess,
 	BgWriterProcess,
+	ArchiverProcess,
 	CheckpointerProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
@@ -429,6 +430,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBootstrapProcess()		(MyAuxProcType == BootstrapProcess)
 #define AmStartupProcess()			(MyAuxProcType == StartupProcess)
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
+#define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
diff --git a/src/include/postmaster/pgarch.h b/src/include/postmaster/pgarch.h
index d102a21ab7..640692d33e 100644
--- a/src/include/postmaster/pgarch.h
+++ b/src/include/postmaster/pgarch.h
@@ -26,14 +26,10 @@
 #define MAX_XFN_CHARS	40
 #define VALID_XFN_CHARS "0123456789ABCDEF.history.backup.partial"
 
-/* ----------
- * Functions called from postmaster
- * ----------
- */
-extern int	pgarch_start(void);
-
-#ifdef EXEC_BACKEND
-extern void PgArchiverMain(int argc, char *argv[]) pg_attribute_noreturn();
-#endif
+extern Size PgArchShmemSize(void);
+extern void PgArchShmemInit(void);
+extern bool PgArchIsSuppressed(void);
+extern void PgArchiverMain(void) pg_attribute_noreturn();
+extern void PgArchWakeup(void);
 
 #endif							/* _PGARCH_H */
diff --git a/src/include/storage/pmsignal.h b/src/include/storage/pmsignal.h
index dbbed18f61..8ed4d87ae6 100644
--- a/src/include/storage/pmsignal.h
+++ b/src/include/storage/pmsignal.h
@@ -34,7 +34,6 @@ typedef enum
 {
 	PMSIGNAL_RECOVERY_STARTED,	/* recovery has started */
 	PMSIGNAL_BEGIN_HOT_STANDBY, /* begin Hot Standby */
-	PMSIGNAL_WAKEN_ARCHIVER,	/* send a NOTIFY signal to xlog archiver */
 	PMSIGNAL_ROTATE_LOGFILE,	/* send SIGUSR1 to syslogger to rotate logfile */
 	PMSIGNAL_START_AUTOVAC_LAUNCHER,	/* start an autovacuum launcher */
 	PMSIGNAL_START_AUTOVAC_WORKER,	/* start an autovacuum worker */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a777cb64a1..2fd1ff09a7 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -370,11 +370,11 @@ extern PGPROC *PreparedXactProcs;
  * We set aside some extra PGPROC structures for auxiliary processes,
  * ie things that aren't full-fledged backends but need shmem access.
  *
- * Background writer, checkpointer and WAL writer run during normal operation.
- * Startup process and WAL receiver also consume 2 slots, but WAL writer is
- * launched only after startup has exited, so we only need 4 slots.
+ * Background writer, checkpointer, WAL writer and archiver run during normal
+ * operation.  Startup process and WAL receiver also consume 2 slots, but WAL
+ * writer is launched only after startup has exited, so we only need 5 slots.
  */
-#define NUM_AUXILIARY_PROCS		4
+#define NUM_AUXILIARY_PROCS		5
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 99a34be465..78e852569e 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1572,6 +1572,7 @@ PGresAttValue
 PGresParamDesc
 PGresult
 PGresult_data
+PgArchData
 PHANDLE
 PLAINTREE
 PLUID_AND_ATTRIBUTES
-- 
2.27.0


----Next_Part(Fri_Mar_12_13_49_06_2021_682)----





^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* Re: Add system identifier to backup manifest
@ 2024-02-15 09:35 Amul Sul <[email protected]>
  2024-02-15 12:11 ` Re: Add system identifier to backup manifest Robert Haas <[email protected]>
  0 siblings, 1 reply; 281+ messages in thread

From: Amul Sul @ 2024-02-15 09:35 UTC (permalink / raw)
  To: Michael Paquier <[email protected]>; +Cc: Robert Haas <[email protected]>; pgsql-hackers

On Thu, Feb 15, 2024 at 7:18 AM Michael Paquier <[email protected]> wrote:

> On Wed, Feb 14, 2024 at 12:29:07PM +0530, Amul Sul wrote:
> > Ok, I did that way in the attached version, I have passed the control
> file's
> > full path as a second argument to verify_system_identifier() what we
> gets in
> > verify_backup_file(), but that is not doing any useful things with it,
> > since we
> > were using get_controlfile() to open the control file, which takes the
> > directory as an input and computes the full path on its own.
>
> I've read through the patch, and that's pretty cool.
>

Thank you for looking into this.


> -static void
> -parse_manifest_file(char *manifest_path, manifest_files_hash **ht_p,
> -                                       manifest_wal_range
> **first_wal_range_p)
> +static manifest_data *
> +parse_manifest_file(char *manifest_path)
>
> In 0001, should the comment describing this routine be updated as
> well?
>

Ok, updated in the attached version.


>
> +   identifier with pg_control of the backup directory or fails
> verification
>
> This is missing a <filename> markup here.
>

Done, in the attached version.


>
> +      <productname>PostgreSQL</productname> 17, it is 2; in older
> versions,
> +      it is 1.
>
> Perhaps a couple of <literal>s here.
>
Done.


> +       if (strcmp(parse->manifest_version, "1") != 0 &&
> +               strcmp(parse->manifest_version, "2") != 0)
> +               json_manifest_parse_failure(parse->context,
> +
>  "unexpected manifest version");
> +
> +       /* Parse version. */
> +       version = strtoi64(parse->manifest_version, &ep, 10);
> +       if (*ep)
> +               json_manifest_parse_failure(parse->context,
> +
>  "manifest version not an integer");
> +
> +       /* Invoke the callback for version */
> +       context->version_cb(context, version);
>
> Shouldn't these two checks be reversed?  And is there actually a need
> for the first check at all knowing that the version callback should be
> in charge of performing the validation vased on the version received?
>

Make sense, reversed the order.

I think, particular allowed versions should be placed at the central place,
and
the callback can check and react on the versions suitable to them, IMHO.


> +my $node2;
> +{
> +       local $ENV{'INITDB_TEMPLATE'} = undef;
>
> Not sure that it is a good idea to duplicate this pattern twice.
> Shouldn't this be something we'd want to control with an option in the
> init() method instead?
>

Yes, I did that in a separate patch, see 0001 patch.


> +static void
> +verify_system_identifier(verifier_context *context, char *controlpath)
>
> Relying both on controlpath, being a full path to the control file
> including the data directory, and context->backup_directory to read
> the contents of the control file looks a bit weird.  Wouldn't it be
> cleaner to just use one of them?
>

Well, yes, I had to have the same feeling, how about having another function
that can accept a full path of pg_control?

I tried in the 0002 patch, where the original function is renamed to
get_dir_controlfile(), which accepts the data directory path as before, and
get_controlfile() now accepts the full path of the pg_control file.

Kindly have a look at the attached version.

Regards,
Amul


Attachments:

  [application/octet-stream] v7-0004-Add-system-identifier-to-backup-manifest.patch (31.0K, ../../CAAJ_b961KftvFLJYvJ1-CWk13dLVcb0+o=q0BRd3n8NJ-ceWBQ@mail.gmail.com/3-v7-0004-Add-system-identifier-to-backup-manifest.patch)
  download | inline diff:
From 19372deb1cb62f685b646548820605d9357561a6 Mon Sep 17 00:00:00 2001
From: Amul Sul <[email protected]>
Date: Thu, 15 Feb 2024 14:42:28 +0530
Subject: [PATCH v7 4/4] Add system identifier to backup manifest

The backup manifest will be having a new item as "System-Identifier"
and its value is from "ControlFileData.system_identifier" when
manifest generated.  This help identify the correct database server
and/or backup while taking subsequent backup.
---
 doc/src/sgml/backup-manifest.sgml             | 16 +++-
 doc/src/sgml/ref/pg_verifybackup.sgml         |  7 +-
 src/backend/backup/backup_manifest.c          |  7 +-
 src/backend/backup/basebackup_incremental.c   | 39 +++++++++
 src/bin/pg_basebackup/t/010_pg_basebackup.pl  | 16 ++++
 src/bin/pg_combinebackup/load_manifest.c      | 33 +++++++-
 src/bin/pg_combinebackup/load_manifest.h      |  1 +
 src/bin/pg_combinebackup/pg_combinebackup.c   | 36 ++++++--
 src/bin/pg_combinebackup/t/005_integrity.pl   | 14 ++++
 src/bin/pg_combinebackup/write_manifest.c     |  8 +-
 src/bin/pg_combinebackup/write_manifest.h     |  3 +-
 src/bin/pg_verifybackup/pg_verifybackup.c     | 82 ++++++++++++++++++-
 src/bin/pg_verifybackup/t/003_corruption.pl   | 26 +++++-
 src/bin/pg_verifybackup/t/005_bad_manifest.pl | 17 +++-
 src/common/parse_manifest.c                   | 82 ++++++++++++++++++-
 src/include/common/parse_manifest.h           |  6 ++
 16 files changed, 368 insertions(+), 25 deletions(-)

diff --git a/doc/src/sgml/backup-manifest.sgml b/doc/src/sgml/backup-manifest.sgml
index 771be1310a1..c43af36dc6f 100644
--- a/doc/src/sgml/backup-manifest.sgml
+++ b/doc/src/sgml/backup-manifest.sgml
@@ -37,7 +37,21 @@
     <term><literal>PostgreSQL-Backup-Manifest-Version</literal></term>
     <listitem>
      <para>
-      The associated value is always the integer 1.
+      The associated value is an integer. Beginning in
+      <productname>PostgreSQL</productname> <literal>17</literal>,
+      it is <literal>2</literal>; in older versions, it is <literal>1</literal>.
+     </para>
+    </listitem>
+   </varlistentry>
+
+   <varlistentry>
+    <term><literal>System-Identifier</literal></term>
+    <listitem>
+     <para>
+      The associated integer value is an unique system identifier to ensure
+      file match up with the installation that produced them. Available only
+      when <literal>PostgreSQL-Backup-Manifest-Version</literal> is
+      <literal>2</literal> and later.
      </para>
     </listitem>
    </varlistentry>
diff --git a/doc/src/sgml/ref/pg_verifybackup.sgml b/doc/src/sgml/ref/pg_verifybackup.sgml
index 36335e0a188..a3f167f9f6e 100644
--- a/doc/src/sgml/ref/pg_verifybackup.sgml
+++ b/doc/src/sgml/ref/pg_verifybackup.sgml
@@ -53,9 +53,10 @@ PostgreSQL documentation
    Backup verification proceeds in four stages. First,
    <literal>pg_verifybackup</literal> reads the
    <literal>backup_manifest</literal> file. If that file
-   does not exist, cannot be read, is malformed, or fails verification
-   against its own internal checksum, <literal>pg_verifybackup</literal>
-   will terminate with a fatal error.
+   does not exist, cannot be read, is malformed, fails to match the system
+   identifier with <filename>pg_control</filename> of the backup directory or
+   fails verification against its own internal checksum,
+   <literal>pg_verifybackup</literal> will terminate with a fatal error.
   </para>
 
   <para>
diff --git a/src/backend/backup/backup_manifest.c b/src/backend/backup/backup_manifest.c
index 2c34e597523..0f31ae72e78 100644
--- a/src/backend/backup/backup_manifest.c
+++ b/src/backend/backup/backup_manifest.c
@@ -13,6 +13,7 @@
 #include "postgres.h"
 
 #include "access/timeline.h"
+#include "access/xlog.h"
 #include "backup/backup_manifest.h"
 #include "backup/basebackup_sink.h"
 #include "libpq/libpq.h"
@@ -79,8 +80,10 @@ InitializeBackupManifest(backup_manifest_info *manifest,
 
 	if (want_manifest != MANIFEST_OPTION_NO)
 		AppendToManifest(manifest,
-						 "{ \"PostgreSQL-Backup-Manifest-Version\": 1,\n"
-						 "\"Files\": [");
+						 "{ \"PostgreSQL-Backup-Manifest-Version\": 2,\n"
+						 "\"System-Identifier\": " UINT64_FORMAT ",\n"
+						 "\"Files\": [",
+						 GetSystemIdentifier());
 }
 
 /*
diff --git a/src/backend/backup/basebackup_incremental.c b/src/backend/backup/basebackup_incremental.c
index 0504c465db8..46e7d90438e 100644
--- a/src/backend/backup/basebackup_incremental.c
+++ b/src/backend/backup/basebackup_incremental.c
@@ -112,6 +112,10 @@ struct IncrementalBackupInfo
 	BlockRefTable *brtab;
 };
 
+static void manifest_process_version(JsonManifestParseContext *context,
+									 int manifest_version);
+static void manifest_process_system_identifier(JsonManifestParseContext *context,
+											   uint64 manifest_system_identifier);
 static void manifest_process_file(JsonManifestParseContext *context,
 								  char *pathname,
 								  size_t size,
@@ -198,6 +202,8 @@ FinalizeIncrementalManifest(IncrementalBackupInfo *ib)
 
 	/* Parse the manifest. */
 	context.private_data = ib;
+	context.version_cb = manifest_process_version;
+	context.system_identifier_cb = manifest_process_system_identifier;
 	context.per_file_cb = manifest_process_file;
 	context.per_wal_range_cb = manifest_process_wal_range;
 	context.error_cb = manifest_report_error;
@@ -910,6 +916,39 @@ hash_string_pointer(const char *s)
 	return hash_bytes(ss, strlen(s));
 }
 
+/*
+ * This callback to validate the manifest version for incremental backup.
+ */
+static void
+manifest_process_version(JsonManifestParseContext *context,
+						 int manifest_version)
+{
+	/* Incremental backups supported on manifest version 2 or later */
+	if (manifest_version == 1)
+		context->error_cb(context,
+						  "backup manifest version 1 does not support incremental backup");
+}
+
+/*
+ * This callback to validate the manifest system identifier against the current
+ * database server.
+ */
+static void
+manifest_process_system_identifier(JsonManifestParseContext *context,
+								   uint64 manifest_system_identifier)
+{
+	uint64		system_identifier;
+
+	/* Get system identifier of current system */
+	system_identifier = GetSystemIdentifier();
+
+	if (manifest_system_identifier != system_identifier)
+		context->error_cb(context,
+						  "manifest system identifier is %llu, but database system identifier is %llu",
+						  (unsigned long long) manifest_system_identifier,
+						  (unsigned long long) system_identifier);
+}
+
 /*
  * This callback is invoked for each file mentioned in the backup manifest.
  *
diff --git a/src/bin/pg_basebackup/t/010_pg_basebackup.pl b/src/bin/pg_basebackup/t/010_pg_basebackup.pl
index 86cc01a640b..e4592834852 100644
--- a/src/bin/pg_basebackup/t/010_pg_basebackup.pl
+++ b/src/bin/pg_basebackup/t/010_pg_basebackup.pl
@@ -966,4 +966,20 @@ $backupdir = $node->backup_dir . '/backup3';
 my @dst_tblspc = glob "$backupdir/pg_tblspc/$tblspc_oid/PG_*";
 is(@dst_tblspc, 1, 'tblspc directory copied');
 
+# Can't take backup with referring manifest of different cluster
+#
+# Set up another new database instance with force initdb option. We don't want
+# to initializing database system by copying initdb template for this, because
+# we want it to be a separate cluster with a different system ID.
+my $node2 = PostgreSQL::Test::Cluster->new('node2');
+$node2->init(force_initdb => 1, has_archiving => 1, allows_streaming => 1);
+$node2->append_conf('postgresql.conf', 'summarize_wal = on');
+$node2->start;
+
+$node2->command_fails_like(
+	[ @pg_basebackup_defs, '-D', "$tempdir" . '/diff_sysid',
+		'--incremental', "$backupdir" . '/backup_manifest' ],
+	qr/manifest system identifier is .*, but database system identifier is/,
+	"pg_basebackup fails with different database system manifest");
+
 done_testing();
diff --git a/src/bin/pg_combinebackup/load_manifest.c b/src/bin/pg_combinebackup/load_manifest.c
index 2b8e74fcf38..4eae7f61efc 100644
--- a/src/bin/pg_combinebackup/load_manifest.c
+++ b/src/bin/pg_combinebackup/load_manifest.c
@@ -50,6 +50,10 @@ static uint32 hash_string_pointer(char *s);
 #define SH_DEFINE
 #include "lib/simplehash.h"
 
+static void combinebackup_version_cb(JsonManifestParseContext *context,
+									 int manifest_version);
+static void combinebackup_system_identifier_cb(JsonManifestParseContext *context,
+											   uint64 manifest_system_identifier);
 static void combinebackup_per_file_cb(JsonManifestParseContext *context,
 									  char *pathname, size_t size,
 									  pg_checksum_type checksum_type,
@@ -103,8 +107,8 @@ load_backup_manifest(char *backup_directory)
 	manifest_files_hash *ht;
 	char	   *buffer;
 	int			rc;
-	JsonManifestParseContext context;
 	manifest_data *result;
+	JsonManifestParseContext context;
 
 	/* Open the manifest file. */
 	snprintf(pathname, MAXPGPATH, "%s/backup_manifest", backup_directory);
@@ -153,6 +157,8 @@ load_backup_manifest(char *backup_directory)
 	result = pg_malloc0(sizeof(manifest_data));
 	result->files = ht;
 	context.private_data = result;
+	context.version_cb = combinebackup_version_cb;
+	context.system_identifier_cb = combinebackup_system_identifier_cb;
 	context.per_file_cb = combinebackup_per_file_cb;
 	context.per_wal_range_cb = combinebackup_per_wal_range_cb;
 	context.error_cb = report_manifest_error;
@@ -181,6 +187,31 @@ report_manifest_error(JsonManifestParseContext *context, const char *fmt,...)
 	exit(1);
 }
 
+/*
+ * This callback to validate the manifest version number for incremental backup.
+ */
+static void
+combinebackup_version_cb(JsonManifestParseContext *context,
+						 int manifest_version)
+{
+	/* Incremental backups supported on manifest version 2 or later */
+	if (manifest_version == 1)
+		pg_fatal("backup manifest version 1 does not support incremental backup");
+}
+
+/*
+ * Record system identifier extracted from the backup manifest.
+ */
+static void
+combinebackup_system_identifier_cb(JsonManifestParseContext *context,
+								   uint64 manifest_system_identifier)
+{
+	manifest_data *manifest = context->private_data;
+
+	/* Validation will be at the later stage */
+	manifest->system_identifier = manifest_system_identifier;
+}
+
 /*
  * Record details extracted from the backup manifest for one file.
  */
diff --git a/src/bin/pg_combinebackup/load_manifest.h b/src/bin/pg_combinebackup/load_manifest.h
index 9163e071afd..8a5a70e4477 100644
--- a/src/bin/pg_combinebackup/load_manifest.h
+++ b/src/bin/pg_combinebackup/load_manifest.h
@@ -55,6 +55,7 @@ typedef struct manifest_wal_range
  */
 typedef struct manifest_data
 {
+	uint64		system_identifier;
 	manifest_files_hash *files;
 	manifest_wal_range *first_wal_range;
 	manifest_wal_range *last_wal_range;
diff --git a/src/bin/pg_combinebackup/pg_combinebackup.c b/src/bin/pg_combinebackup/pg_combinebackup.c
index 31ead7f4058..f214a72a5ad 100644
--- a/src/bin/pg_combinebackup/pg_combinebackup.c
+++ b/src/bin/pg_combinebackup/pg_combinebackup.c
@@ -92,7 +92,7 @@ cb_cleanup_dir *cleanup_dir_list = NULL;
 
 static void add_tablespace_mapping(cb_options *opt, char *arg);
 static StringInfo check_backup_label_files(int n_backups, char **backup_dirs);
-static void check_control_files(int n_backups, char **backup_dirs);
+static uint64 check_control_files(int n_backups, char **backup_dirs);
 static void check_input_dir_permissions(char *dir);
 static void cleanup_directories_atexit(void);
 static void create_output_directory(char *dirname, cb_options *opt);
@@ -134,11 +134,13 @@ main(int argc, char *argv[])
 
 	const char *progname;
 	char	   *last_input_dir;
+	int			i;
 	int			optindex;
 	int			c;
 	int			n_backups;
 	int			n_prior_backups;
 	int			version;
+	uint64		system_identifier;
 	char	  **prior_backup_dirs;
 	cb_options	opt;
 	cb_tablespace *tablespaces;
@@ -216,7 +218,7 @@ main(int argc, char *argv[])
 
 	/* Sanity-check control files. */
 	n_backups = argc - optind;
-	check_control_files(n_backups, argv + optind);
+	system_identifier = check_control_files(n_backups, argv + optind);
 
 	/* Sanity-check backup_label files, and get the contents of the last one. */
 	last_backup_label = check_backup_label_files(n_backups, argv + optind);
@@ -231,6 +233,26 @@ main(int argc, char *argv[])
 	/* Load backup manifests. */
 	manifests = load_backup_manifests(n_backups, prior_backup_dirs);
 
+	/*
+	 * Validate the manifest system identifier against the backup system
+	 * identifier.
+	 */
+	for (i = 0; i < n_backups; i++)
+	{
+		if (manifests[i] &&
+			manifests[i]->system_identifier != system_identifier)
+		{
+			char	   *controlpath;
+
+			controlpath = psprintf("%s/%s", prior_backup_dirs[i], "global/pg_control");
+
+			pg_fatal("%s: manifest system identifier is %llu, but control file has %llu",
+					 controlpath,
+					 (unsigned long long) manifests[i]->system_identifier,
+					 (unsigned long long) system_identifier);
+		}
+	}
+
 	/* Figure out which tablespaces are going to be included in the output. */
 	last_input_dir = argv[argc - 1];
 	check_input_dir_permissions(last_input_dir);
@@ -256,7 +278,7 @@ main(int argc, char *argv[])
 	/* If we need to write a backup_manifest, prepare to do so. */
 	if (!opt.dry_run && !opt.no_manifest)
 	{
-		mwriter = create_manifest_writer(opt.output);
+		mwriter = create_manifest_writer(opt.output, system_identifier);
 
 		/*
 		 * Verify that we have a backup manifest for the final backup; else we
@@ -517,9 +539,9 @@ check_backup_label_files(int n_backups, char **backup_dirs)
 }
 
 /*
- * Sanity check control files.
+ * Sanity check control files and return system_identifier.
  */
-static void
+static uint64
 check_control_files(int n_backups, char **backup_dirs)
 {
 	int			i;
@@ -534,7 +556,7 @@ check_control_files(int n_backups, char **backup_dirs)
 
 		controlpath = psprintf("%s/%s", backup_dirs[i], "global/pg_control");
 		pg_log_debug("reading \"%s\"", controlpath);
-		control_file = get_controlfile(backup_dirs[i], &crc_ok);
+		control_file = get_dir_controlfile(backup_dirs[i], &crc_ok);
 
 		/* Control file contents not meaningful if CRC is bad. */
 		if (!crc_ok)
@@ -564,6 +586,8 @@ check_control_files(int n_backups, char **backup_dirs)
 	 */
 	pg_log_debug("system identifier is %llu",
 				 (unsigned long long) system_identifier);
+
+	return system_identifier;
 }
 
 /*
diff --git a/src/bin/pg_combinebackup/t/005_integrity.pl b/src/bin/pg_combinebackup/t/005_integrity.pl
index 5d425209211..d79348ab6f3 100644
--- a/src/bin/pg_combinebackup/t/005_integrity.pl
+++ b/src/bin/pg_combinebackup/t/005_integrity.pl
@@ -8,6 +8,7 @@ use strict;
 use warnings FATAL => 'all';
 use File::Compare;
 use File::Path qw(rmtree);
+use File::Copy;
 use PostgreSQL::Test::Cluster;
 use PostgreSQL::Test::Utils;
 use Test::More;
@@ -80,6 +81,19 @@ $node1->command_fails_like(
 	qr/expected system identifier.*but found/,
 	"can't combine backups from different nodes");
 
+# Can't combine when different manifest system identifier
+rename("$backup2path/backup_manifest", "$backup2path/backup_manifest.orig")
+  or BAIL_OUT "could not move $backup2path/backup_manifest";
+copy("$backupother2path/backup_manifest", "$backup2path/backup_manifest")
+  or BAIL_OUT "could not copy $backupother2path/backup_manifest";
+$node1->command_fails_like(
+	[ 'pg_combinebackup', $backup1path, $backup2path, $backup3path, '-o', $resultpath ],
+	qr/ manifest system identifier is .*, but control file has /,
+	"can't combine backups with different manifest system identifier ");
+# Restore the backup state
+move("$backup2path/backup_manifest.orig", "$backup2path/backup_manifest")
+  or BAIL_OUT "could not move $backup2path/backup_manifest";
+
 # Can't omit a required backup.
 $node1->command_fails_like(
 	[ 'pg_combinebackup', $backup1path, $backup3path, '-o', $resultpath ],
diff --git a/src/bin/pg_combinebackup/write_manifest.c b/src/bin/pg_combinebackup/write_manifest.c
index 01deb82cc9f..7a2065e1db7 100644
--- a/src/bin/pg_combinebackup/write_manifest.c
+++ b/src/bin/pg_combinebackup/write_manifest.c
@@ -45,7 +45,7 @@ static size_t hex_encode(const uint8 *src, size_t len, char *dst);
  * in the specified directory.
  */
 manifest_writer *
-create_manifest_writer(char *directory)
+create_manifest_writer(char *directory, uint64 system_identifier)
 {
 	manifest_writer *mwriter = pg_malloc(sizeof(manifest_writer));
 
@@ -57,8 +57,10 @@ create_manifest_writer(char *directory)
 	pg_checksum_init(&mwriter->manifest_ctx, CHECKSUM_TYPE_SHA256);
 
 	appendStringInfo(&mwriter->buf,
-					 "{ \"PostgreSQL-Backup-Manifest-Version\": 1,\n"
-					 "\"Files\": [");
+					 "{ \"PostgreSQL-Backup-Manifest-Version\": 2,\n"
+					 "\"System-Identifier\": " UINT64_FORMAT ",\n"
+					 "\"Files\": [",
+					 system_identifier);
 
 	return mwriter;
 }
diff --git a/src/bin/pg_combinebackup/write_manifest.h b/src/bin/pg_combinebackup/write_manifest.h
index de0f742779f..ebc4f9441ad 100644
--- a/src/bin/pg_combinebackup/write_manifest.h
+++ b/src/bin/pg_combinebackup/write_manifest.h
@@ -19,7 +19,8 @@ struct manifest_wal_range;
 struct manifest_writer;
 typedef struct manifest_writer manifest_writer;
 
-extern manifest_writer *create_manifest_writer(char *directory);
+extern manifest_writer *create_manifest_writer(char *directory,
+											   uint64 system_identifier);
 extern void add_file_to_manifest(manifest_writer *mwriter,
 								 const char *manifest_path,
 								 size_t size, time_t mtime,
diff --git a/src/bin/pg_verifybackup/pg_verifybackup.c b/src/bin/pg_verifybackup/pg_verifybackup.c
index 8561678a7d0..7f730514f3e 100644
--- a/src/bin/pg_verifybackup/pg_verifybackup.c
+++ b/src/bin/pg_verifybackup/pg_verifybackup.c
@@ -18,6 +18,7 @@
 #include <sys/stat.h>
 #include <time.h>
 
+#include "common/controldata_utils.h"
 #include "common/hashfn.h"
 #include "common/logging.h"
 #include "common/parse_manifest.h"
@@ -98,6 +99,8 @@ typedef struct manifest_wal_range
  */
 typedef struct manifest_data
 {
+	int			version;
+	uint64		system_identifier;
 	manifest_files_hash *files;
 	manifest_wal_range *first_wal_range;
 	manifest_wal_range *last_wal_range;
@@ -116,6 +119,10 @@ typedef struct verifier_context
 } verifier_context;
 
 static manifest_data *parse_manifest_file(char *manifest_path);
+static void verifybackup_version_cb(JsonManifestParseContext *context,
+									int manifest_version);
+static void verifybackup_system_identifier(JsonManifestParseContext *context,
+										   uint64 manifest_system_identifier);
 static void verifybackup_per_file_cb(JsonManifestParseContext *context,
 									 char *pathname, size_t size,
 									 pg_checksum_type checksum_type,
@@ -129,6 +136,8 @@ static void report_manifest_error(JsonManifestParseContext *context,
 								  const char *fmt,...)
 			pg_attribute_printf(2, 3) pg_attribute_noreturn();
 
+static void verify_system_identifier(manifest_data *manifest,
+									 char *controlpath);
 static void verify_backup_directory(verifier_context *context,
 									char *relpath, char *fullpath);
 static void verify_backup_file(verifier_context *context,
@@ -375,9 +384,7 @@ main(int argc, char **argv)
 }
 
 /*
- * Parse a manifest file. Construct a hash table with information about
- * all the files it mentions, and a linked list of all the WAL ranges it
- * mentions.
+ * Parse a manifest file and return a data structure describing the contents.
  */
 static manifest_data *
 parse_manifest_file(char *manifest_path)
@@ -432,6 +439,8 @@ parse_manifest_file(char *manifest_path)
 	result = pg_malloc0(sizeof(manifest_data));
 	result->files = ht;
 	context.private_data = result;
+	context.version_cb = verifybackup_version_cb;
+	context.system_identifier_cb = verifybackup_system_identifier;
 	context.per_file_cb = verifybackup_per_file_cb;
 	context.per_wal_range_cb = verifybackup_per_wal_range_cb;
 	context.error_cb = report_manifest_error;
@@ -461,6 +470,32 @@ report_manifest_error(JsonManifestParseContext *context, const char *fmt,...)
 	exit(1);
 }
 
+/*
+ * Record details extracted from the backup manifest.
+ */
+static void
+verifybackup_version_cb(JsonManifestParseContext *context,
+						int manifest_version)
+{
+	manifest_data *manifest = context->private_data;
+
+	/* Validation will be at the later stage */
+	manifest->version = manifest_version;
+}
+
+/*
+ * Record details extracted from the backup manifest.
+ */
+static void
+verifybackup_system_identifier(JsonManifestParseContext *context,
+						 uint64 manifest_system_identifier)
+{
+	manifest_data *manifest = context->private_data;
+
+	/* Validation will be at the later stage */
+	manifest->system_identifier = manifest_system_identifier;
+}
+
 /*
  * Record details extracted from the backup manifest for one file.
  */
@@ -517,6 +552,39 @@ verifybackup_per_wal_range_cb(JsonManifestParseContext *context,
 	manifest->last_wal_range = range;
 }
 
+/*
+ * Sanity check control file of the backup and validate system identifier
+ * against manifest system identifier.
+ */
+static void
+verify_system_identifier(manifest_data *manifest, char *controlpath)
+{
+	ControlFileData *control_file;
+	bool		crc_ok;
+
+	pg_log_debug("reading \"%s\"", controlpath);
+	control_file = get_controlfile(controlpath, &crc_ok);
+
+	/* Control file contents not meaningful if CRC is bad. */
+	if (!crc_ok)
+		report_fatal_error("%s: CRC is incorrect", controlpath);
+
+	/* Can't interpret control file if not current version. */
+	if (control_file->pg_control_version != PG_CONTROL_VERSION)
+		report_fatal_error("%s: unexpected control file version",
+						   controlpath);
+
+	/* System identifiers should match. */
+	if (manifest->system_identifier != control_file->system_identifier)
+		report_fatal_error("%s: manifest system identifier is %llu, but control file has %llu",
+						   controlpath,
+						   (unsigned long long) manifest->system_identifier,
+						   (unsigned long long) control_file->system_identifier);
+
+	/* Release memory. */
+	pfree(control_file);
+}
+
 /*
  * Verify one directory.
  *
@@ -650,6 +718,14 @@ verify_backup_file(verifier_context *context, char *relpath, char *fullpath)
 		m->bad = true;
 	}
 
+	/*
+	 * Validate the manifest system identifier, not available in manifest
+	 * version 1.
+	 */
+	if (context->manifest->version != 1 &&
+		strcmp(relpath, "global/pg_control") == 0)
+			verify_system_identifier(context->manifest, fullpath);
+
 	/* Update statistics for progress report, if necessary */
 	if (show_progress && !skip_checksums && should_verify_checksum(m))
 		total_size += m->size;
diff --git a/src/bin/pg_verifybackup/t/003_corruption.pl b/src/bin/pg_verifybackup/t/003_corruption.pl
index 11bd5770818..36d032288fe 100644
--- a/src/bin/pg_verifybackup/t/003_corruption.pl
+++ b/src/bin/pg_verifybackup/t/003_corruption.pl
@@ -6,6 +6,7 @@
 use strict;
 use warnings FATAL => 'all';
 use File::Path qw(rmtree);
+use File::Copy;
 use PostgreSQL::Test::Cluster;
 use PostgreSQL::Test::Utils;
 use Test::More;
@@ -68,6 +69,11 @@ my @scenario = (
 		'mutilate' => \&mutilate_replace_file,
 		'fails_like' => qr/checksum mismatch for file/
 	},
+	{
+		'name' => 'system_identifier',
+		'mutilate' => \&mutilate_system_identifier,
+		'fails_like' => qr/manifest system identifier is .*, but control file has/
+	},
 	{
 		'name' => 'bad_manifest',
 		'mutilate' => \&mutilate_bad_manifest,
@@ -216,7 +222,7 @@ sub mutilate_append_to_file
 sub mutilate_truncate_file
 {
 	my ($backup_path) = @_;
-	my $pathname = "$backup_path/global/pg_control";
+	my $pathname = "$backup_path/pg_hba.conf";
 	open(my $fh, '>', $pathname) || die "open $pathname: $!";
 	close($fh);
 	return;
@@ -236,6 +242,24 @@ sub mutilate_replace_file
 	return;
 }
 
+# Copy manifest of other backups to demonstrate the case where the wrong
+# manifest is referred
+sub mutilate_system_identifier
+{
+	my ($backup_path) = @_;
+
+	# Set up another new database instance with different system identifier and
+	# make backup
+	my $node = PostgreSQL::Test::Cluster->new('node');
+	$node->init(force_initdb => 1, allows_streaming => 1);
+	$node->start;
+	$node->backup('backup2');
+	move($node->backup_dir.'/backup2/backup_manifest', $backup_path.'/backup_manifest')
+		or BAIL_OUT "could not copy manifest to $backup_path";
+	$node->teardown_node(fail_ok => 1);
+	return;
+}
+
 # Corrupt the backup manifest.
 sub mutilate_bad_manifest
 {
diff --git a/src/bin/pg_verifybackup/t/005_bad_manifest.pl b/src/bin/pg_verifybackup/t/005_bad_manifest.pl
index e278ccea5a2..f95b45a7230 100644
--- a/src/bin/pg_verifybackup/t/005_bad_manifest.pl
+++ b/src/bin/pg_verifybackup/t/005_bad_manifest.pl
@@ -29,10 +29,14 @@ test_parse_error('expected version indicator', <<EOM);
 {"not-expected": 1}
 EOM
 
-test_parse_error('unexpected manifest version', <<EOM);
+test_parse_error('manifest version not an integer', <<EOM);
 {"PostgreSQL-Backup-Manifest-Version": "phooey"}
 EOM
 
+test_parse_error('unexpected manifest version', <<EOM);
+{"PostgreSQL-Backup-Manifest-Version": 9876599}
+EOM
+
 test_parse_error('unexpected scalar', <<EOM);
 {"PostgreSQL-Backup-Manifest-Version": 1, "Files": true}
 EOM
@@ -156,6 +160,17 @@ test_parse_error('expected at least 2 lines', <<EOM);
 {"PostgreSQL-Backup-Manifest-Version": 1, "Files": [], "Manifest-Checksum": null}
 EOM
 
+test_parse_error('unrecognized top-level field', <<EOM);
+{"PostgreSQL-Backup-Manifest-Version": 1,
+ "System-Identifier": "7320280665284567118"
+ "Manifest-Checksum": "eabdc9caf8a" }
+EOM
+
+test_parse_error('expected system identifier', <<EOM);
+{"PostgreSQL-Backup-Manifest-Version": 2,
+ "Manifest-Checksum": "eabdc9caf8a" }
+EOM
+
 my $manifest_without_newline = <<EOM;
 {"PostgreSQL-Backup-Manifest-Version": 1,
  "Files": [],
diff --git a/src/common/parse_manifest.c b/src/common/parse_manifest.c
index 92a97714f36..7e3682f75b0 100644
--- a/src/common/parse_manifest.c
+++ b/src/common/parse_manifest.c
@@ -25,6 +25,7 @@ typedef enum
 	JM_EXPECT_TOPLEVEL_END,
 	JM_EXPECT_TOPLEVEL_FIELD,
 	JM_EXPECT_VERSION_VALUE,
+	JM_EXPECT_SYSTEM_IDENTIFIER_VALUE,
 	JM_EXPECT_FILES_START,
 	JM_EXPECT_FILES_NEXT,
 	JM_EXPECT_THIS_FILE_FIELD,
@@ -85,6 +86,9 @@ typedef struct
 
 	/* Miscellaneous other stuff. */
 	bool		saw_version_field;
+	bool		saw_system_identifier_field;
+	char	   *manifest_version;
+	char	   *manifest_system_identifier;
 	char	   *manifest_checksum;
 } JsonManifestParseState;
 
@@ -96,6 +100,8 @@ static JsonParseErrorType json_manifest_object_field_start(void *state, char *fn
 														   bool isnull);
 static JsonParseErrorType json_manifest_scalar(void *state, char *token,
 											   JsonTokenType tokentype);
+static void json_manifest_finalize_version(JsonManifestParseState *parse);
+static void json_manifest_finalize_system_identifier(JsonManifestParseState *parse);
 static void json_manifest_finalize_file(JsonManifestParseState *parse);
 static void json_manifest_finalize_wal_range(JsonManifestParseState *parse);
 static void verify_manifest_checksum(JsonManifestParseState *parse,
@@ -128,6 +134,7 @@ json_parse_manifest(JsonManifestParseContext *context, char *buffer,
 	parse.context = context;
 	parse.state = JM_EXPECT_TOPLEVEL_START;
 	parse.saw_version_field = false;
+	parse.saw_system_identifier_field = false;
 
 	/* Create a JSON lexing context. */
 	lex = makeJsonLexContextCstringLen(NULL, buffer, size, PG_UTF8, true);
@@ -311,6 +318,16 @@ json_manifest_object_field_start(void *state, char *fname, bool isnull)
 				parse->saw_version_field = true;
 				break;
 			}
+			else if (!parse->saw_system_identifier_field &&
+					 strcmp(parse->manifest_version, "1") != 0)
+			{
+				if (strcmp(fname, "System-Identifier") != 0)
+					json_manifest_parse_failure(parse->context,
+												"expected system identifier");
+				parse->state = JM_EXPECT_SYSTEM_IDENTIFIER_VALUE;
+				parse->saw_system_identifier_field = true;
+				break;
+			}
 
 			/* Is this the list of files? */
 			if (strcmp(fname, "Files") == 0)
@@ -404,9 +421,14 @@ json_manifest_scalar(void *state, char *token, JsonTokenType tokentype)
 	switch (parse->state)
 	{
 		case JM_EXPECT_VERSION_VALUE:
-			if (strcmp(token, "1") != 0)
-				json_manifest_parse_failure(parse->context,
-											"unexpected manifest version");
+			parse->manifest_version = token;
+			json_manifest_finalize_version(parse);
+			parse->state = JM_EXPECT_TOPLEVEL_FIELD;
+			break;
+
+		case JM_EXPECT_SYSTEM_IDENTIFIER_VALUE:
+			parse->manifest_system_identifier = token;
+			json_manifest_finalize_system_identifier(parse);
 			parse->state = JM_EXPECT_TOPLEVEL_FIELD;
 			break;
 
@@ -464,6 +486,60 @@ json_manifest_scalar(void *state, char *token, JsonTokenType tokentype)
 	return JSON_SUCCESS;
 }
 
+/*
+ * Do additional parsing and sanity-checking of the manifest version, and invoke
+ * the callback so that the caller can gets that detail and take actions
+ * accordingly.  This happens for each manifest when the corresponding JSON
+ * object is completely parsed.
+ */
+static void
+json_manifest_finalize_version(JsonManifestParseState *parse)
+{
+	JsonManifestParseContext *context = parse->context;
+	int			version;
+	char	   *ep;
+
+	Assert(parse->saw_version_field);
+
+	/* Parse version. */
+	version = strtoi64(parse->manifest_version, &ep, 10);
+	if (*ep)
+		json_manifest_parse_failure(parse->context,
+									"manifest version not an integer");
+
+	if (version != 1 && version != 2)
+		json_manifest_parse_failure(parse->context,
+									"unexpected manifest version");
+
+	/* Invoke the callback for version */
+	context->version_cb(context, version);
+}
+
+/*
+ * Do additional parsing and sanity-checking of the system identifier, and
+ * invoke the callback so that the caller can gets that detail and take actions
+ * accordingly.  This happens for each manifest when the corresponding JSON
+ * object is completely parsed.
+ */
+static void
+json_manifest_finalize_system_identifier(JsonManifestParseState *parse)
+{
+	JsonManifestParseContext *context = parse->context;
+	uint64		system_identifier;
+	char	   *ep;
+
+	Assert(parse->saw_system_identifier_field);
+
+	/* Parse system identifier. */
+	system_identifier = strtou64(parse->manifest_system_identifier, &ep, 10);
+	if (*ep)
+		json_manifest_parse_failure(parse->context,
+									"manifest system identifier not an integer");
+
+	/* Invoke the callback for system identifier */
+	context->system_identifier_cb(context, system_identifier);
+}
+
 /*
  * Do additional parsing and sanity-checking of the details gathered for one
  * file, and invoke the per-file callback so that the caller gets those
diff --git a/src/include/common/parse_manifest.h b/src/include/common/parse_manifest.h
index f74be0db35f..78b052c045b 100644
--- a/src/include/common/parse_manifest.h
+++ b/src/include/common/parse_manifest.h
@@ -21,6 +21,10 @@
 struct JsonManifestParseContext;
 typedef struct JsonManifestParseContext JsonManifestParseContext;
 
+typedef void (*json_manifest_version_callback) (JsonManifestParseContext *,
+												int manifest_version);
+typedef void (*json_manifest_system_identifier_callback) (JsonManifestParseContext *,
+														  uint64 manifest_system_identifier);
 typedef void (*json_manifest_per_file_callback) (JsonManifestParseContext *,
 												 char *pathname,
 												 size_t size, pg_checksum_type checksum_type,
@@ -35,6 +39,8 @@ typedef void (*json_manifest_error_callback) (JsonManifestParseContext *,
 struct JsonManifestParseContext
 {
 	void	   *private_data;
+	json_manifest_version_callback version_cb;
+	json_manifest_system_identifier_callback system_identifier_cb;
 	json_manifest_per_file_callback per_file_cb;
 	json_manifest_per_wal_range_callback per_wal_range_cb;
 	json_manifest_error_callback error_cb;
-- 
2.18.0



  [application/octet-stream] v7-0003-pg_verifybackup-code-refactor.patch (8.0K, ../../CAAJ_b961KftvFLJYvJ1-CWk13dLVcb0+o=q0BRd3n8NJ-ceWBQ@mail.gmail.com/4-v7-0003-pg_verifybackup-code-refactor.patch)
  download | inline diff:
From e955751ee1c54e72091fcecca68624a8d3cf7cf9 Mon Sep 17 00:00:00 2001
From: Amul Sul <[email protected]>
Date: Mon, 22 Jan 2024 10:45:19 +0530
Subject: [PATCH v7 3/4] pg_verifybackup: code refactor

Rename parser_context to manifest_data and make parse_manifest_file
return that instead of out-argument.
---
 src/bin/pg_verifybackup/pg_verifybackup.c | 75 ++++++++++-------------
 1 file changed, 34 insertions(+), 41 deletions(-)

diff --git a/src/bin/pg_verifybackup/pg_verifybackup.c b/src/bin/pg_verifybackup/pg_verifybackup.c
index ae8c18f3737..8561678a7d0 100644
--- a/src/bin/pg_verifybackup/pg_verifybackup.c
+++ b/src/bin/pg_verifybackup/pg_verifybackup.c
@@ -94,31 +94,28 @@ typedef struct manifest_wal_range
 } manifest_wal_range;
 
 /*
- * Details we need in callbacks that occur while parsing a backup manifest.
+ * All the data parsed from a backup_manifest file.
  */
-typedef struct parser_context
+typedef struct manifest_data
 {
-	manifest_files_hash *ht;
+	manifest_files_hash *files;
 	manifest_wal_range *first_wal_range;
 	manifest_wal_range *last_wal_range;
-} parser_context;
+} manifest_data;
 
 /*
  * All of the context information we need while checking a backup manifest.
  */
 typedef struct verifier_context
 {
-	manifest_files_hash *ht;
+	manifest_data *manifest;
 	char	   *backup_directory;
 	SimpleStringList ignore_list;
 	bool		exit_on_error;
 	bool		saw_any_error;
 } verifier_context;
 
-static void parse_manifest_file(char *manifest_path,
-								manifest_files_hash **ht_p,
-								manifest_wal_range **first_wal_range_p);
-
+static manifest_data *parse_manifest_file(char *manifest_path);
 static void verifybackup_per_file_cb(JsonManifestParseContext *context,
 									 char *pathname, size_t size,
 									 pg_checksum_type checksum_type,
@@ -142,8 +139,7 @@ static void verify_file_checksum(verifier_context *context,
 								 manifest_file *m, char *fullpath);
 static void parse_required_wal(verifier_context *context,
 							   char *pg_waldump_path,
-							   char *wal_directory,
-							   manifest_wal_range *first_wal_range);
+							   char *wal_directory);
 
 static void report_backup_error(verifier_context *context,
 								const char *pg_restrict fmt,...)
@@ -185,7 +181,6 @@ main(int argc, char **argv)
 
 	int			c;
 	verifier_context context;
-	manifest_wal_range *first_wal_range;
 	char	   *manifest_path = NULL;
 	bool		no_parse_wal = false;
 	bool		quiet = false;
@@ -338,7 +333,7 @@ main(int argc, char **argv)
 	 * the manifest as fatal; there doesn't seem to be much point in trying to
 	 * verify the backup directory against a corrupted manifest.
 	 */
-	parse_manifest_file(manifest_path, &context.ht, &first_wal_range);
+	context.manifest = parse_manifest_file(manifest_path);
 
 	/*
 	 * Now scan the files in the backup directory. At this stage, we verify
@@ -367,8 +362,7 @@ main(int argc, char **argv)
 	 * not to do so.
 	 */
 	if (!no_parse_wal)
-		parse_required_wal(&context, pg_waldump_path,
-						   wal_directory, first_wal_range);
+		parse_required_wal(&context, pg_waldump_path, wal_directory);
 
 	/*
 	 * If everything looks OK, tell the user this, unless we were asked to
@@ -385,9 +379,8 @@ main(int argc, char **argv)
  * all the files it mentions, and a linked list of all the WAL ranges it
  * mentions.
  */
-static void
-parse_manifest_file(char *manifest_path, manifest_files_hash **ht_p,
-					manifest_wal_range **first_wal_range_p)
+static manifest_data *
+parse_manifest_file(char *manifest_path)
 {
 	int			fd;
 	struct stat statbuf;
@@ -396,8 +389,8 @@ parse_manifest_file(char *manifest_path, manifest_files_hash **ht_p,
 	manifest_files_hash *ht;
 	char	   *buffer;
 	int			rc;
-	parser_context private_context;
 	JsonManifestParseContext context;
+	manifest_data *result;
 
 	/* Open the manifest file. */
 	if ((fd = open(manifest_path, O_RDONLY | PG_BINARY, 0)) < 0)
@@ -436,10 +429,9 @@ parse_manifest_file(char *manifest_path, manifest_files_hash **ht_p,
 	close(fd);
 
 	/* Parse the manifest. */
-	private_context.ht = ht;
-	private_context.first_wal_range = NULL;
-	private_context.last_wal_range = NULL;
-	context.private_data = &private_context;
+	result = pg_malloc0(sizeof(manifest_data));
+	result->files = ht;
+	context.private_data = result;
 	context.per_file_cb = verifybackup_per_file_cb;
 	context.per_wal_range_cb = verifybackup_per_wal_range_cb;
 	context.error_cb = report_manifest_error;
@@ -448,9 +440,7 @@ parse_manifest_file(char *manifest_path, manifest_files_hash **ht_p,
 	/* Done with the buffer. */
 	pfree(buffer);
 
-	/* Return the file hash table and WAL range list we constructed. */
-	*ht_p = ht;
-	*first_wal_range_p = private_context.first_wal_range;
+	return result;
 }
 
 /*
@@ -480,8 +470,8 @@ verifybackup_per_file_cb(JsonManifestParseContext *context,
 						 pg_checksum_type checksum_type,
 						 int checksum_length, uint8 *checksum_payload)
 {
-	parser_context *pcxt = context->private_data;
-	manifest_files_hash *ht = pcxt->ht;
+	manifest_data *manifest = context->private_data;
+	manifest_files_hash *ht = manifest->files;
 	manifest_file *m;
 	bool		found;
 
@@ -508,7 +498,7 @@ verifybackup_per_wal_range_cb(JsonManifestParseContext *context,
 							  TimeLineID tli,
 							  XLogRecPtr start_lsn, XLogRecPtr end_lsn)
 {
-	parser_context *pcxt = context->private_data;
+	manifest_data *manifest = context->private_data;
 	manifest_wal_range *range;
 
 	/* Allocate and initialize a struct describing this WAL range. */
@@ -516,15 +506,15 @@ verifybackup_per_wal_range_cb(JsonManifestParseContext *context,
 	range->tli = tli;
 	range->start_lsn = start_lsn;
 	range->end_lsn = end_lsn;
-	range->prev = pcxt->last_wal_range;
+	range->prev = manifest->last_wal_range;
 	range->next = NULL;
 
 	/* Add it to the end of the list. */
-	if (pcxt->first_wal_range == NULL)
-		pcxt->first_wal_range = range;
+	if (manifest->first_wal_range == NULL)
+		manifest->first_wal_range = range;
 	else
-		pcxt->last_wal_range->next = range;
-	pcxt->last_wal_range = range;
+		manifest->last_wal_range->next = range;
+	manifest->last_wal_range = range;
 }
 
 /*
@@ -639,7 +629,7 @@ verify_backup_file(verifier_context *context, char *relpath, char *fullpath)
 	}
 
 	/* Check whether there's an entry in the manifest hash. */
-	m = manifest_files_lookup(context->ht, relpath);
+	m = manifest_files_lookup(context->manifest->files, relpath);
 	if (m == NULL)
 	{
 		report_backup_error(context,
@@ -679,11 +669,12 @@ verify_backup_file(verifier_context *context, char *relpath, char *fullpath)
 static void
 report_extra_backup_files(verifier_context *context)
 {
+	manifest_data *manifest = context->manifest;
 	manifest_files_iterator it;
 	manifest_file *m;
 
-	manifest_files_start_iterate(context->ht, &it);
-	while ((m = manifest_files_iterate(context->ht, &it)) != NULL)
+	manifest_files_start_iterate(manifest->files, &it);
+	while ((m = manifest_files_iterate(manifest->files, &it)) != NULL)
 		if (!m->matched && !should_ignore_relpath(context, m->pathname))
 			report_backup_error(context,
 								"\"%s\" is present in the manifest but not on disk",
@@ -698,13 +689,14 @@ report_extra_backup_files(verifier_context *context)
 static void
 verify_backup_checksums(verifier_context *context)
 {
+	manifest_data *manifest = context->manifest;
 	manifest_files_iterator it;
 	manifest_file *m;
 
 	progress_report(false);
 
-	manifest_files_start_iterate(context->ht, &it);
-	while ((m = manifest_files_iterate(context->ht, &it)) != NULL)
+	manifest_files_start_iterate(manifest->files, &it);
+	while ((m = manifest_files_iterate(manifest->files, &it)) != NULL)
 	{
 		if (should_verify_checksum(m) &&
 			!should_ignore_relpath(context, m->pathname))
@@ -833,9 +825,10 @@ verify_file_checksum(verifier_context *context, manifest_file *m,
  */
 static void
 parse_required_wal(verifier_context *context, char *pg_waldump_path,
-				   char *wal_directory, manifest_wal_range *first_wal_range)
+				   char *wal_directory)
 {
-	manifest_wal_range *this_wal_range = first_wal_range;
+	manifest_data *manifest = context->manifest;
+	manifest_wal_range *this_wal_range = manifest->first_wal_range;
 
 	while (this_wal_range != NULL)
 	{
-- 
2.18.0



  [application/octet-stream] v7-0001-Add-option-to-force-initdb-in-PostgreSQL-Test-Clu.patch (3.3K, ../../CAAJ_b961KftvFLJYvJ1-CWk13dLVcb0+o=q0BRd3n8NJ-ceWBQ@mail.gmail.com/5-v7-0001-Add-option-to-force-initdb-in-PostgreSQL-Test-Clu.patch)
  download | inline diff:
From e859d940df23ebc482ca81e130e1d55b12447051 Mon Sep 17 00:00:00 2001
From: Amul Sul <[email protected]>
Date: Thu, 15 Feb 2024 12:10:23 +0530
Subject: [PATCH v7 1/4] Add option to force initdb in
 PostgreSQL::Test::Cluster:init()

---
 src/bin/pg_combinebackup/t/005_integrity.pl | 19 +++++++------------
 src/test/perl/PostgreSQL/Test/Cluster.pm    | 12 +++++++-----
 2 files changed, 14 insertions(+), 17 deletions(-)

diff --git a/src/bin/pg_combinebackup/t/005_integrity.pl b/src/bin/pg_combinebackup/t/005_integrity.pl
index 3b445d0e30f..5d425209211 100644
--- a/src/bin/pg_combinebackup/t/005_integrity.pl
+++ b/src/bin/pg_combinebackup/t/005_integrity.pl
@@ -18,18 +18,13 @@ $node1->init(has_archiving => 1, allows_streaming => 1);
 $node1->append_conf('postgresql.conf', 'summarize_wal = on');
 $node1->start;
 
-# Set up another new database instance. We don't want to use the cached
-# INITDB_TEMPLATE for this, because we want it to be a separate cluster
-# with a different system ID.
-my $node2;
-{
-	local $ENV{'INITDB_TEMPLATE'} = undef;
-
-	$node2 = PostgreSQL::Test::Cluster->new('node2');
-	$node2->init(has_archiving => 1, allows_streaming => 1);
-	$node2->append_conf('postgresql.conf', 'summarize_wal = on');
-	$node2->start;
-}
+# Set up another new database instance with force initdb option. We don't want
+# to initializing database system by copying initdb template for this, because
+# we want it to be a separate cluster with a different system ID.
+my $node2 = PostgreSQL::Test::Cluster->new('node2');
+$node2->init(force_initdb => 1, has_archiving => 1, allows_streaming => 1);
+$node2->append_conf('postgresql.conf', 'summarize_wal = on');
+$node2->start;
 
 # Take a full backup from node1.
 my $backup1path = $node1->backup_dir . '/backup1';
diff --git a/src/test/perl/PostgreSQL/Test/Cluster.pm b/src/test/perl/PostgreSQL/Test/Cluster.pm
index e2e70d0dbf9..136a29d0977 100644
--- a/src/test/perl/PostgreSQL/Test/Cluster.pm
+++ b/src/test/perl/PostgreSQL/Test/Cluster.pm
@@ -518,18 +518,20 @@ sub init
 
 	$params{allows_streaming} = 0 unless defined $params{allows_streaming};
 	$params{has_archiving} = 0 unless defined $params{has_archiving};
+	$params{force_initdb} = 0 unless defined $params{force_initdb};
 
 	mkdir $self->backup_dir;
 	mkdir $self->archive_dir;
 
-	# If available and if there aren't any parameters, use a previously
-	# initdb'd cluster as a template by copying it. For a lot of tests, that's
-	# substantially cheaper. Do so only if there aren't parameters, it doesn't
-	# seem worth figuring out whether they affect compatibility.
+	# If available and if there aren't any parameters, use a previously initdb'd
+	# cluster as a template by copying it unless asked initdb explicitly. For a
+	# lot of tests, that's substantially cheaper. Do so only if there aren't
+	# parameters, it doesn't seem worth figuring out whether they affect
+	# compatibility.
 	#
 	# There's very similar code in pg_regress.c, but we can't easily
 	# deduplicate it until we require perl at build time.
-	if (defined $params{extra} or !defined $ENV{INITDB_TEMPLATE})
+	if ($params{force_initdb} or defined $params{extra} or !defined $ENV{INITDB_TEMPLATE})
 	{
 		note("initializing database system by running initdb");
 		PostgreSQL::Test::Utils::system_or_bail('initdb', '-D', $pgdata, '-A',
-- 
2.18.0



  [application/octet-stream] v7-0002-Code-refactor-get_controlfile-to-accept-full-path.patch (5.4K, ../../CAAJ_b961KftvFLJYvJ1-CWk13dLVcb0+o=q0BRd3n8NJ-ceWBQ@mail.gmail.com/6-v7-0002-Code-refactor-get_controlfile-to-accept-full-path.patch)
  download | inline diff:
From fff5c6a6f912e2822e07381fa57dc32082475efc Mon Sep 17 00:00:00 2001
From: Amul Sul <[email protected]>
Date: Thu, 15 Feb 2024 14:33:12 +0530
Subject: [PATCH v7 2/4] Code refactor: get_controlfile() to accept full path
 of control file

The current version get_controlfile() accepts data directory path, and
computes the full path for the pg_control file, but it would be better
to have a version that accepts the full path of that pg_control file.
---
 src/backend/utils/misc/pg_controldata.c |  8 ++++----
 src/bin/pg_checksums/pg_checksums.c     |  2 +-
 src/bin/pg_controldata/pg_controldata.c |  2 +-
 src/bin/pg_ctl/pg_ctl.c                 |  2 +-
 src/common/controldata_utils.c          | 19 ++++++++++++++++---
 src/include/common/controldata_utils.h  |  3 ++-
 6 files changed, 25 insertions(+), 11 deletions(-)

diff --git a/src/backend/utils/misc/pg_controldata.c b/src/backend/utils/misc/pg_controldata.c
index 55435dbcf3a..ea8eb3624d9 100644
--- a/src/backend/utils/misc/pg_controldata.c
+++ b/src/backend/utils/misc/pg_controldata.c
@@ -44,7 +44,7 @@ pg_control_system(PG_FUNCTION_ARGS)
 
 	/* read the control file */
 	LWLockAcquire(ControlFileLock, LW_SHARED);
-	ControlFile = get_controlfile(DataDir, &crc_ok);
+	ControlFile = get_dir_controlfile(DataDir, &crc_ok);
 	LWLockRelease(ControlFileLock);
 	if (!crc_ok)
 		ereport(ERROR,
@@ -84,7 +84,7 @@ pg_control_checkpoint(PG_FUNCTION_ARGS)
 
 	/* Read the control file. */
 	LWLockAcquire(ControlFileLock, LW_SHARED);
-	ControlFile = get_controlfile(DataDir, &crc_ok);
+	ControlFile = get_dir_controlfile(DataDir, &crc_ok);
 	LWLockRelease(ControlFileLock);
 	if (!crc_ok)
 		ereport(ERROR,
@@ -175,7 +175,7 @@ pg_control_recovery(PG_FUNCTION_ARGS)
 
 	/* read the control file */
 	LWLockAcquire(ControlFileLock, LW_SHARED);
-	ControlFile = get_controlfile(DataDir, &crc_ok);
+	ControlFile = get_dir_controlfile(DataDir, &crc_ok);
 	LWLockRelease(ControlFileLock);
 	if (!crc_ok)
 		ereport(ERROR,
@@ -216,7 +216,7 @@ pg_control_init(PG_FUNCTION_ARGS)
 
 	/* read the control file */
 	LWLockAcquire(ControlFileLock, LW_SHARED);
-	ControlFile = get_controlfile(DataDir, &crc_ok);
+	ControlFile = get_dir_controlfile(DataDir, &crc_ok);
 	LWLockRelease(ControlFileLock);
 	if (!crc_ok)
 		ereport(ERROR,
diff --git a/src/bin/pg_checksums/pg_checksums.c b/src/bin/pg_checksums/pg_checksums.c
index 9e6fd435f60..4bdd85f42e0 100644
--- a/src/bin/pg_checksums/pg_checksums.c
+++ b/src/bin/pg_checksums/pg_checksums.c
@@ -545,7 +545,7 @@ main(int argc, char *argv[])
 	}
 
 	/* Read the control file and check compatibility */
-	ControlFile = get_controlfile(DataDir, &crc_ok);
+	ControlFile = get_dir_controlfile(DataDir, &crc_ok);
 	if (!crc_ok)
 		pg_fatal("pg_control CRC value is incorrect");
 
diff --git a/src/bin/pg_controldata/pg_controldata.c b/src/bin/pg_controldata/pg_controldata.c
index 93e0837947c..1e615131612 100644
--- a/src/bin/pg_controldata/pg_controldata.c
+++ b/src/bin/pg_controldata/pg_controldata.c
@@ -165,7 +165,7 @@ main(int argc, char *argv[])
 	}
 
 	/* get a copy of the control file */
-	ControlFile = get_controlfile(DataDir, &crc_ok);
+	ControlFile = get_dir_controlfile(DataDir, &crc_ok);
 	if (!crc_ok)
 	{
 		pg_log_warning("calculated CRC checksum does not match value stored in control file");
diff --git a/src/bin/pg_ctl/pg_ctl.c b/src/bin/pg_ctl/pg_ctl.c
index 6900b27675e..1d887c1b9df 100644
--- a/src/bin/pg_ctl/pg_ctl.c
+++ b/src/bin/pg_ctl/pg_ctl.c
@@ -2174,7 +2174,7 @@ get_control_dbstate(void)
 {
 	DBState		ret;
 	bool		crc_ok;
-	ControlFileData *control_file_data = get_controlfile(pg_data, &crc_ok);
+	ControlFileData *control_file_data = get_dir_controlfile(pg_data, &crc_ok);
 
 	if (!crc_ok)
 	{
diff --git a/src/common/controldata_utils.c b/src/common/controldata_utils.c
index 92e8fed6b2e..43566920fed 100644
--- a/src/common/controldata_utils.c
+++ b/src/common/controldata_utils.c
@@ -49,11 +49,10 @@
  * file data is correct.
  */
 ControlFileData *
-get_controlfile(const char *DataDir, bool *crc_ok_p)
+get_controlfile(const char *ControlFilePath, bool *crc_ok_p)
 {
 	ControlFileData *ControlFile;
 	int			fd;
-	char		ControlFilePath[MAXPGPATH];
 	pg_crc32c	crc;
 	int			r;
 #ifdef FRONTEND
@@ -64,7 +63,6 @@ get_controlfile(const char *DataDir, bool *crc_ok_p)
 	Assert(crc_ok_p);
 
 	ControlFile = palloc_object(ControlFileData);
-	snprintf(ControlFilePath, MAXPGPATH, "%s/global/pg_control", DataDir);
 
 #ifdef FRONTEND
 	INIT_CRC32C(last_crc);
@@ -162,6 +160,21 @@ retry:
 	return ControlFile;
 }
 
+/*
+ * get_dir_controlfile()
+ *
+ * Get controlfile values of the given data directory.
+ */
+ControlFileData *
+get_dir_controlfile(const char *DataDir, bool *crc_ok_p)
+{
+	char		ControlFilePath[MAXPGPATH];
+
+	snprintf(ControlFilePath, MAXPGPATH, "%s/global/pg_control", DataDir);
+
+	return get_controlfile(ControlFilePath, crc_ok_p);
+}
+
 /*
  * update_controlfile()
  *
diff --git a/src/include/common/controldata_utils.h b/src/include/common/controldata_utils.h
index 04da70e87b2..56528360663 100644
--- a/src/include/common/controldata_utils.h
+++ b/src/include/common/controldata_utils.h
@@ -12,7 +12,8 @@
 
 #include "catalog/pg_control.h"
 
-extern ControlFileData *get_controlfile(const char *DataDir, bool *crc_ok_p);
+extern ControlFileData *get_controlfile(const char *ControlFilePath, bool *crc_ok_p);
+extern ControlFileData *get_dir_controlfile(const char *DataDir, bool *crc_ok_p);
 extern void update_controlfile(const char *DataDir,
 							   ControlFileData *ControlFile, bool do_sync);
 
-- 
2.18.0



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* Re: Add system identifier to backup manifest
  2024-02-15 09:35 Re: Add system identifier to backup manifest Amul Sul <[email protected]>
@ 2024-02-15 12:11 ` Robert Haas <[email protected]>
  2024-02-18 22:52   ` Re: Add system identifier to backup manifest Michael Paquier <[email protected]>
  0 siblings, 1 reply; 281+ messages in thread

From: Robert Haas @ 2024-02-15 12:11 UTC (permalink / raw)
  To: Amul Sul <[email protected]>; +Cc: Michael Paquier <[email protected]>; pgsql-hackers

On Thu, Feb 15, 2024 at 3:05 PM Amul Sul <[email protected]> wrote:
> Kindly have a look at the attached version.

IMHO, 0001 looks fine, except probably the comment could be phrased a
bit more nicely. That can be left for whoever commits this to
wordsmith. Michael, what are your plans?

0002 seems like a reasonable approach, but there's a hunk in the wrong
patch: 0004 modifies pg_combinebackup's check_control_files to use
get_dir_controlfile() rather than git_controlfile(), but it looks to
me like that should be part of 0002.

-- 
Robert Haas
EDB: http://www.enterprisedb.com






^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* Re: Add system identifier to backup manifest
  2024-02-15 09:35 Re: Add system identifier to backup manifest Amul Sul <[email protected]>
  2024-02-15 12:11 ` Re: Add system identifier to backup manifest Robert Haas <[email protected]>
@ 2024-02-18 22:52   ` Michael Paquier <[email protected]>
  2024-02-19 06:36     ` Re: Add system identifier to backup manifest Amul Sul <[email protected]>
  0 siblings, 1 reply; 281+ messages in thread

From: Michael Paquier @ 2024-02-18 22:52 UTC (permalink / raw)
  To: Robert Haas <[email protected]>; +Cc: Amul Sul <[email protected]>; pgsql-hackers

On Thu, Feb 15, 2024 at 05:41:46PM +0530, Robert Haas wrote:
> On Thu, Feb 15, 2024 at 3:05 PM Amul Sul <[email protected]> wrote:
> > Kindly have a look at the attached version.
> 
> IMHO, 0001 looks fine, except probably the comment could be phrased a
> bit more nicely.

And the new option should be documented at the top of the init()
routine for perldoc.

> That can be left for whoever commits this to
> wordsmith. Michael, what are your plans?

Not much, so feel free to not wait for me.  I've just read through the
patch because I like the idea/feature :)

> 0002 seems like a reasonable approach, but there's a hunk in the wrong
> patch: 0004 modifies pg_combinebackup's check_control_files to use
> get_dir_controlfile() rather than git_controlfile(), but it looks to
> me like that should be part of 0002.

I'm slightly concerned about 0002 that silently changes the meaning of
get_controlfile().  That would impact extension code without people
knowing about it when compiling, just when they run their stuff under
17~.
--
Michael


Attachments:

  [application/pgp-signature] signature.asc (833B, ../../[email protected]/2-signature.asc)
  download

^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* Re: Add system identifier to backup manifest
  2024-02-15 09:35 Re: Add system identifier to backup manifest Amul Sul <[email protected]>
  2024-02-15 12:11 ` Re: Add system identifier to backup manifest Robert Haas <[email protected]>
  2024-02-18 22:52   ` Re: Add system identifier to backup manifest Michael Paquier <[email protected]>
@ 2024-02-19 06:36     ` Amul Sul <[email protected]>
  2024-02-21 04:31       ` Re: Add system identifier to backup manifest Michael Paquier <[email protected]>
  2024-03-01 05:58       ` Re: Add system identifier to backup manifest Michael Paquier <[email protected]>
  0 siblings, 2 replies; 281+ messages in thread

From: Amul Sul @ 2024-02-19 06:36 UTC (permalink / raw)
  To: Michael Paquier <[email protected]>; +Cc: Robert Haas <[email protected]>; pgsql-hackers

On Mon, Feb 19, 2024 at 4:22 AM Michael Paquier <[email protected]> wrote:

> On Thu, Feb 15, 2024 at 05:41:46PM +0530, Robert Haas wrote:
> > On Thu, Feb 15, 2024 at 3:05 PM Amul Sul <[email protected]> wrote:
> > > Kindly have a look at the attached version.
> >
> > IMHO, 0001 looks fine, except probably the comment could be phrased a
> > bit more nicely.
>
> And the new option should be documented at the top of the init()
> routine for perldoc.
>

Added in the attached version.


> > That can be left for whoever commits this to
> > wordsmith. Michael, what are your plans?
>
> Not much, so feel free to not wait for me.  I've just read through the
> patch because I like the idea/feature :)
>

Thank you, that helped a lot.

> 0002 seems like a reasonable approach, but there's a hunk in the wrong
> > patch: 0004 modifies pg_combinebackup's check_control_files to use
> > get_dir_controlfile() rather than git_controlfile(), but it looks to
> > me like that should be part of 0002.
>

Fixed in the attached version.


> I'm slightly concerned about 0002 that silently changes the meaning of
> get_controlfile().  That would impact extension code without people
> knowing about it when compiling, just when they run their stuff under
> 17~.
>

Agreed, now they will have an error as _could not read file "<DataDir>": Is
a
directory_. But, IIUC, that what usually happens with the dev version, and
the
extension needs to be updated for compatibility with the newer version for
the
same reason.

Regards,
Amul


Attachments:

  [application/x-patch] v8-0001-Add-option-to-force-initdb-in-PostgreSQL-Test-Clu.patch (3.8K, ../../CAAJ_b97_9nJCD3kAmNfSnN1s7m9n1h97OP=APwts=g-rWGztxQ@mail.gmail.com/3-v8-0001-Add-option-to-force-initdb-in-PostgreSQL-Test-Clu.patch)
  download | inline diff:
From 354014538b16579f005bd6f4ce771b1aa22b5e02 Mon Sep 17 00:00:00 2001
From: Amul Sul <[email protected]>
Date: Thu, 15 Feb 2024 12:10:23 +0530
Subject: [PATCH v8 1/4] Add option to force initdb in
 PostgreSQL::Test::Cluster:init()

---
 src/bin/pg_combinebackup/t/005_integrity.pl | 19 +++++++------------
 src/test/perl/PostgreSQL/Test/Cluster.pm    | 15 ++++++++++-----
 2 files changed, 17 insertions(+), 17 deletions(-)

diff --git a/src/bin/pg_combinebackup/t/005_integrity.pl b/src/bin/pg_combinebackup/t/005_integrity.pl
index 3b445d0e30f..5d425209211 100644
--- a/src/bin/pg_combinebackup/t/005_integrity.pl
+++ b/src/bin/pg_combinebackup/t/005_integrity.pl
@@ -18,18 +18,13 @@ $node1->init(has_archiving => 1, allows_streaming => 1);
 $node1->append_conf('postgresql.conf', 'summarize_wal = on');
 $node1->start;
 
-# Set up another new database instance. We don't want to use the cached
-# INITDB_TEMPLATE for this, because we want it to be a separate cluster
-# with a different system ID.
-my $node2;
-{
-	local $ENV{'INITDB_TEMPLATE'} = undef;
-
-	$node2 = PostgreSQL::Test::Cluster->new('node2');
-	$node2->init(has_archiving => 1, allows_streaming => 1);
-	$node2->append_conf('postgresql.conf', 'summarize_wal = on');
-	$node2->start;
-}
+# Set up another new database instance with force initdb option. We don't want
+# to initializing database system by copying initdb template for this, because
+# we want it to be a separate cluster with a different system ID.
+my $node2 = PostgreSQL::Test::Cluster->new('node2');
+$node2->init(force_initdb => 1, has_archiving => 1, allows_streaming => 1);
+$node2->append_conf('postgresql.conf', 'summarize_wal = on');
+$node2->start;
 
 # Take a full backup from node1.
 my $backup1path = $node1->backup_dir . '/backup1';
diff --git a/src/test/perl/PostgreSQL/Test/Cluster.pm b/src/test/perl/PostgreSQL/Test/Cluster.pm
index 07da74cf562..2b4f9a48365 100644
--- a/src/test/perl/PostgreSQL/Test/Cluster.pm
+++ b/src/test/perl/PostgreSQL/Test/Cluster.pm
@@ -495,6 +495,10 @@ a directory that's only accessible to the current user to ensure that.
 On Windows, we use SSPI authentication to ensure the same (by pg_regress
 --config-auth).
 
+force_initdb => 1 will force to initialized the cluster by initdb. Otherwise, if
+available and if there aren't any parameters, use a previously initdb'd cluster
+as a template by copying it.
+
 WAL archiving can be enabled on this node by passing the keyword parameter
 has_archiving => 1. This is disabled by default.
 
@@ -517,6 +521,7 @@ sub init
 
 	local %ENV = $self->_get_env();
 
+	$params{force_initdb} = 0 unless defined $params{force_initdb};
 	$params{allows_streaming} = 0 unless defined $params{allows_streaming};
 	$params{has_archiving} = 0 unless defined $params{has_archiving};
 
@@ -529,14 +534,14 @@ sub init
 	mkdir $self->backup_dir;
 	mkdir $self->archive_dir;
 
-	# If available and if there aren't any parameters, use a previously
-	# initdb'd cluster as a template by copying it. For a lot of tests, that's
-	# substantially cheaper. Do so only if there aren't parameters, it doesn't
-	# seem worth figuring out whether they affect compatibility.
+	# For a lot of tests, that's substantially cheaper to copy previously
+	# initdb'd cluster as a template. Do so only if force_initdb => 0, and there
+	# aren't parameters, it doesn't seem worth figuring out whether they affect
+	# compatibility.
 	#
 	# There's very similar code in pg_regress.c, but we can't easily
 	# deduplicate it until we require perl at build time.
-	if (defined $params{extra} or !defined $ENV{INITDB_TEMPLATE})
+	if ($params{force_initdb} or defined $params{extra} or !defined $ENV{INITDB_TEMPLATE})
 	{
 		note("initializing database system by running initdb");
 		PostgreSQL::Test::Utils::system_or_bail('initdb', '-D', $pgdata, '-A',
-- 
2.18.0



  [application/x-patch] v8-0003-pg_verifybackup-code-refactor.patch (8.0K, ../../CAAJ_b97_9nJCD3kAmNfSnN1s7m9n1h97OP=APwts=g-rWGztxQ@mail.gmail.com/4-v8-0003-pg_verifybackup-code-refactor.patch)
  download | inline diff:
From 93d617f3387e5c066541f298e5b71260ebd15103 Mon Sep 17 00:00:00 2001
From: Amul Sul <[email protected]>
Date: Mon, 22 Jan 2024 10:45:19 +0530
Subject: [PATCH v8 3/4] pg_verifybackup: code refactor

Rename parser_context to manifest_data and make parse_manifest_file
return that instead of out-argument.
---
 src/bin/pg_verifybackup/pg_verifybackup.c | 75 ++++++++++-------------
 1 file changed, 34 insertions(+), 41 deletions(-)

diff --git a/src/bin/pg_verifybackup/pg_verifybackup.c b/src/bin/pg_verifybackup/pg_verifybackup.c
index ae8c18f3737..8561678a7d0 100644
--- a/src/bin/pg_verifybackup/pg_verifybackup.c
+++ b/src/bin/pg_verifybackup/pg_verifybackup.c
@@ -94,31 +94,28 @@ typedef struct manifest_wal_range
 } manifest_wal_range;
 
 /*
- * Details we need in callbacks that occur while parsing a backup manifest.
+ * All the data parsed from a backup_manifest file.
  */
-typedef struct parser_context
+typedef struct manifest_data
 {
-	manifest_files_hash *ht;
+	manifest_files_hash *files;
 	manifest_wal_range *first_wal_range;
 	manifest_wal_range *last_wal_range;
-} parser_context;
+} manifest_data;
 
 /*
  * All of the context information we need while checking a backup manifest.
  */
 typedef struct verifier_context
 {
-	manifest_files_hash *ht;
+	manifest_data *manifest;
 	char	   *backup_directory;
 	SimpleStringList ignore_list;
 	bool		exit_on_error;
 	bool		saw_any_error;
 } verifier_context;
 
-static void parse_manifest_file(char *manifest_path,
-								manifest_files_hash **ht_p,
-								manifest_wal_range **first_wal_range_p);
-
+static manifest_data *parse_manifest_file(char *manifest_path);
 static void verifybackup_per_file_cb(JsonManifestParseContext *context,
 									 char *pathname, size_t size,
 									 pg_checksum_type checksum_type,
@@ -142,8 +139,7 @@ static void verify_file_checksum(verifier_context *context,
 								 manifest_file *m, char *fullpath);
 static void parse_required_wal(verifier_context *context,
 							   char *pg_waldump_path,
-							   char *wal_directory,
-							   manifest_wal_range *first_wal_range);
+							   char *wal_directory);
 
 static void report_backup_error(verifier_context *context,
 								const char *pg_restrict fmt,...)
@@ -185,7 +181,6 @@ main(int argc, char **argv)
 
 	int			c;
 	verifier_context context;
-	manifest_wal_range *first_wal_range;
 	char	   *manifest_path = NULL;
 	bool		no_parse_wal = false;
 	bool		quiet = false;
@@ -338,7 +333,7 @@ main(int argc, char **argv)
 	 * the manifest as fatal; there doesn't seem to be much point in trying to
 	 * verify the backup directory against a corrupted manifest.
 	 */
-	parse_manifest_file(manifest_path, &context.ht, &first_wal_range);
+	context.manifest = parse_manifest_file(manifest_path);
 
 	/*
 	 * Now scan the files in the backup directory. At this stage, we verify
@@ -367,8 +362,7 @@ main(int argc, char **argv)
 	 * not to do so.
 	 */
 	if (!no_parse_wal)
-		parse_required_wal(&context, pg_waldump_path,
-						   wal_directory, first_wal_range);
+		parse_required_wal(&context, pg_waldump_path, wal_directory);
 
 	/*
 	 * If everything looks OK, tell the user this, unless we were asked to
@@ -385,9 +379,8 @@ main(int argc, char **argv)
  * all the files it mentions, and a linked list of all the WAL ranges it
  * mentions.
  */
-static void
-parse_manifest_file(char *manifest_path, manifest_files_hash **ht_p,
-					manifest_wal_range **first_wal_range_p)
+static manifest_data *
+parse_manifest_file(char *manifest_path)
 {
 	int			fd;
 	struct stat statbuf;
@@ -396,8 +389,8 @@ parse_manifest_file(char *manifest_path, manifest_files_hash **ht_p,
 	manifest_files_hash *ht;
 	char	   *buffer;
 	int			rc;
-	parser_context private_context;
 	JsonManifestParseContext context;
+	manifest_data *result;
 
 	/* Open the manifest file. */
 	if ((fd = open(manifest_path, O_RDONLY | PG_BINARY, 0)) < 0)
@@ -436,10 +429,9 @@ parse_manifest_file(char *manifest_path, manifest_files_hash **ht_p,
 	close(fd);
 
 	/* Parse the manifest. */
-	private_context.ht = ht;
-	private_context.first_wal_range = NULL;
-	private_context.last_wal_range = NULL;
-	context.private_data = &private_context;
+	result = pg_malloc0(sizeof(manifest_data));
+	result->files = ht;
+	context.private_data = result;
 	context.per_file_cb = verifybackup_per_file_cb;
 	context.per_wal_range_cb = verifybackup_per_wal_range_cb;
 	context.error_cb = report_manifest_error;
@@ -448,9 +440,7 @@ parse_manifest_file(char *manifest_path, manifest_files_hash **ht_p,
 	/* Done with the buffer. */
 	pfree(buffer);
 
-	/* Return the file hash table and WAL range list we constructed. */
-	*ht_p = ht;
-	*first_wal_range_p = private_context.first_wal_range;
+	return result;
 }
 
 /*
@@ -480,8 +470,8 @@ verifybackup_per_file_cb(JsonManifestParseContext *context,
 						 pg_checksum_type checksum_type,
 						 int checksum_length, uint8 *checksum_payload)
 {
-	parser_context *pcxt = context->private_data;
-	manifest_files_hash *ht = pcxt->ht;
+	manifest_data *manifest = context->private_data;
+	manifest_files_hash *ht = manifest->files;
 	manifest_file *m;
 	bool		found;
 
@@ -508,7 +498,7 @@ verifybackup_per_wal_range_cb(JsonManifestParseContext *context,
 							  TimeLineID tli,
 							  XLogRecPtr start_lsn, XLogRecPtr end_lsn)
 {
-	parser_context *pcxt = context->private_data;
+	manifest_data *manifest = context->private_data;
 	manifest_wal_range *range;
 
 	/* Allocate and initialize a struct describing this WAL range. */
@@ -516,15 +506,15 @@ verifybackup_per_wal_range_cb(JsonManifestParseContext *context,
 	range->tli = tli;
 	range->start_lsn = start_lsn;
 	range->end_lsn = end_lsn;
-	range->prev = pcxt->last_wal_range;
+	range->prev = manifest->last_wal_range;
 	range->next = NULL;
 
 	/* Add it to the end of the list. */
-	if (pcxt->first_wal_range == NULL)
-		pcxt->first_wal_range = range;
+	if (manifest->first_wal_range == NULL)
+		manifest->first_wal_range = range;
 	else
-		pcxt->last_wal_range->next = range;
-	pcxt->last_wal_range = range;
+		manifest->last_wal_range->next = range;
+	manifest->last_wal_range = range;
 }
 
 /*
@@ -639,7 +629,7 @@ verify_backup_file(verifier_context *context, char *relpath, char *fullpath)
 	}
 
 	/* Check whether there's an entry in the manifest hash. */
-	m = manifest_files_lookup(context->ht, relpath);
+	m = manifest_files_lookup(context->manifest->files, relpath);
 	if (m == NULL)
 	{
 		report_backup_error(context,
@@ -679,11 +669,12 @@ verify_backup_file(verifier_context *context, char *relpath, char *fullpath)
 static void
 report_extra_backup_files(verifier_context *context)
 {
+	manifest_data *manifest = context->manifest;
 	manifest_files_iterator it;
 	manifest_file *m;
 
-	manifest_files_start_iterate(context->ht, &it);
-	while ((m = manifest_files_iterate(context->ht, &it)) != NULL)
+	manifest_files_start_iterate(manifest->files, &it);
+	while ((m = manifest_files_iterate(manifest->files, &it)) != NULL)
 		if (!m->matched && !should_ignore_relpath(context, m->pathname))
 			report_backup_error(context,
 								"\"%s\" is present in the manifest but not on disk",
@@ -698,13 +689,14 @@ report_extra_backup_files(verifier_context *context)
 static void
 verify_backup_checksums(verifier_context *context)
 {
+	manifest_data *manifest = context->manifest;
 	manifest_files_iterator it;
 	manifest_file *m;
 
 	progress_report(false);
 
-	manifest_files_start_iterate(context->ht, &it);
-	while ((m = manifest_files_iterate(context->ht, &it)) != NULL)
+	manifest_files_start_iterate(manifest->files, &it);
+	while ((m = manifest_files_iterate(manifest->files, &it)) != NULL)
 	{
 		if (should_verify_checksum(m) &&
 			!should_ignore_relpath(context, m->pathname))
@@ -833,9 +825,10 @@ verify_file_checksum(verifier_context *context, manifest_file *m,
  */
 static void
 parse_required_wal(verifier_context *context, char *pg_waldump_path,
-				   char *wal_directory, manifest_wal_range *first_wal_range)
+				   char *wal_directory)
 {
-	manifest_wal_range *this_wal_range = first_wal_range;
+	manifest_data *manifest = context->manifest;
+	manifest_wal_range *this_wal_range = manifest->first_wal_range;
 
 	while (this_wal_range != NULL)
 	{
-- 
2.18.0



  [application/x-patch] v8-0002-Code-refactor-get_controlfile-to-accept-full-path.patch (6.1K, ../../CAAJ_b97_9nJCD3kAmNfSnN1s7m9n1h97OP=APwts=g-rWGztxQ@mail.gmail.com/5-v8-0002-Code-refactor-get_controlfile-to-accept-full-path.patch)
  download | inline diff:
From 788487286a400f0378cd417c90da13be43ad1f18 Mon Sep 17 00:00:00 2001
From: Amul Sul <[email protected]>
Date: Thu, 15 Feb 2024 14:33:12 +0530
Subject: [PATCH v8 2/4] Code refactor: get_controlfile() to accept full path
 of control file

The current version get_controlfile() accepts data directory path, and
computes the full path for the pg_control file, but it would be better
to have a version that accepts the full path of that pg_control file.
---
 src/backend/utils/misc/pg_controldata.c     |  8 ++++----
 src/bin/pg_checksums/pg_checksums.c         |  2 +-
 src/bin/pg_combinebackup/pg_combinebackup.c |  2 +-
 src/bin/pg_controldata/pg_controldata.c     |  2 +-
 src/bin/pg_ctl/pg_ctl.c                     |  2 +-
 src/common/controldata_utils.c              | 19 ++++++++++++++++---
 src/include/common/controldata_utils.h      |  3 ++-
 7 files changed, 26 insertions(+), 12 deletions(-)

diff --git a/src/backend/utils/misc/pg_controldata.c b/src/backend/utils/misc/pg_controldata.c
index 55435dbcf3a..ea8eb3624d9 100644
--- a/src/backend/utils/misc/pg_controldata.c
+++ b/src/backend/utils/misc/pg_controldata.c
@@ -44,7 +44,7 @@ pg_control_system(PG_FUNCTION_ARGS)
 
 	/* read the control file */
 	LWLockAcquire(ControlFileLock, LW_SHARED);
-	ControlFile = get_controlfile(DataDir, &crc_ok);
+	ControlFile = get_dir_controlfile(DataDir, &crc_ok);
 	LWLockRelease(ControlFileLock);
 	if (!crc_ok)
 		ereport(ERROR,
@@ -84,7 +84,7 @@ pg_control_checkpoint(PG_FUNCTION_ARGS)
 
 	/* Read the control file. */
 	LWLockAcquire(ControlFileLock, LW_SHARED);
-	ControlFile = get_controlfile(DataDir, &crc_ok);
+	ControlFile = get_dir_controlfile(DataDir, &crc_ok);
 	LWLockRelease(ControlFileLock);
 	if (!crc_ok)
 		ereport(ERROR,
@@ -175,7 +175,7 @@ pg_control_recovery(PG_FUNCTION_ARGS)
 
 	/* read the control file */
 	LWLockAcquire(ControlFileLock, LW_SHARED);
-	ControlFile = get_controlfile(DataDir, &crc_ok);
+	ControlFile = get_dir_controlfile(DataDir, &crc_ok);
 	LWLockRelease(ControlFileLock);
 	if (!crc_ok)
 		ereport(ERROR,
@@ -216,7 +216,7 @@ pg_control_init(PG_FUNCTION_ARGS)
 
 	/* read the control file */
 	LWLockAcquire(ControlFileLock, LW_SHARED);
-	ControlFile = get_controlfile(DataDir, &crc_ok);
+	ControlFile = get_dir_controlfile(DataDir, &crc_ok);
 	LWLockRelease(ControlFileLock);
 	if (!crc_ok)
 		ereport(ERROR,
diff --git a/src/bin/pg_checksums/pg_checksums.c b/src/bin/pg_checksums/pg_checksums.c
index 9e6fd435f60..4bdd85f42e0 100644
--- a/src/bin/pg_checksums/pg_checksums.c
+++ b/src/bin/pg_checksums/pg_checksums.c
@@ -545,7 +545,7 @@ main(int argc, char *argv[])
 	}
 
 	/* Read the control file and check compatibility */
-	ControlFile = get_controlfile(DataDir, &crc_ok);
+	ControlFile = get_dir_controlfile(DataDir, &crc_ok);
 	if (!crc_ok)
 		pg_fatal("pg_control CRC value is incorrect");
 
diff --git a/src/bin/pg_combinebackup/pg_combinebackup.c b/src/bin/pg_combinebackup/pg_combinebackup.c
index 31ead7f4058..00c018351ac 100644
--- a/src/bin/pg_combinebackup/pg_combinebackup.c
+++ b/src/bin/pg_combinebackup/pg_combinebackup.c
@@ -534,7 +534,7 @@ check_control_files(int n_backups, char **backup_dirs)
 
 		controlpath = psprintf("%s/%s", backup_dirs[i], "global/pg_control");
 		pg_log_debug("reading \"%s\"", controlpath);
-		control_file = get_controlfile(backup_dirs[i], &crc_ok);
+		control_file = get_controlfile(controlpath, &crc_ok);
 
 		/* Control file contents not meaningful if CRC is bad. */
 		if (!crc_ok)
diff --git a/src/bin/pg_controldata/pg_controldata.c b/src/bin/pg_controldata/pg_controldata.c
index 93e0837947c..1e615131612 100644
--- a/src/bin/pg_controldata/pg_controldata.c
+++ b/src/bin/pg_controldata/pg_controldata.c
@@ -165,7 +165,7 @@ main(int argc, char *argv[])
 	}
 
 	/* get a copy of the control file */
-	ControlFile = get_controlfile(DataDir, &crc_ok);
+	ControlFile = get_dir_controlfile(DataDir, &crc_ok);
 	if (!crc_ok)
 	{
 		pg_log_warning("calculated CRC checksum does not match value stored in control file");
diff --git a/src/bin/pg_ctl/pg_ctl.c b/src/bin/pg_ctl/pg_ctl.c
index 6900b27675e..1d887c1b9df 100644
--- a/src/bin/pg_ctl/pg_ctl.c
+++ b/src/bin/pg_ctl/pg_ctl.c
@@ -2174,7 +2174,7 @@ get_control_dbstate(void)
 {
 	DBState		ret;
 	bool		crc_ok;
-	ControlFileData *control_file_data = get_controlfile(pg_data, &crc_ok);
+	ControlFileData *control_file_data = get_dir_controlfile(pg_data, &crc_ok);
 
 	if (!crc_ok)
 	{
diff --git a/src/common/controldata_utils.c b/src/common/controldata_utils.c
index 92e8fed6b2e..43566920fed 100644
--- a/src/common/controldata_utils.c
+++ b/src/common/controldata_utils.c
@@ -49,11 +49,10 @@
  * file data is correct.
  */
 ControlFileData *
-get_controlfile(const char *DataDir, bool *crc_ok_p)
+get_controlfile(const char *ControlFilePath, bool *crc_ok_p)
 {
 	ControlFileData *ControlFile;
 	int			fd;
-	char		ControlFilePath[MAXPGPATH];
 	pg_crc32c	crc;
 	int			r;
 #ifdef FRONTEND
@@ -64,7 +63,6 @@ get_controlfile(const char *DataDir, bool *crc_ok_p)
 	Assert(crc_ok_p);
 
 	ControlFile = palloc_object(ControlFileData);
-	snprintf(ControlFilePath, MAXPGPATH, "%s/global/pg_control", DataDir);
 
 #ifdef FRONTEND
 	INIT_CRC32C(last_crc);
@@ -162,6 +160,21 @@ retry:
 	return ControlFile;
 }
 
+/*
+ * get_dir_controlfile()
+ *
+ * Get controlfile values of the given data directory.
+ */
+ControlFileData *
+get_dir_controlfile(const char *DataDir, bool *crc_ok_p)
+{
+	char		ControlFilePath[MAXPGPATH];
+
+	snprintf(ControlFilePath, MAXPGPATH, "%s/global/pg_control", DataDir);
+
+	return get_controlfile(ControlFilePath, crc_ok_p);
+}
+
 /*
  * update_controlfile()
  *
diff --git a/src/include/common/controldata_utils.h b/src/include/common/controldata_utils.h
index 04da70e87b2..56528360663 100644
--- a/src/include/common/controldata_utils.h
+++ b/src/include/common/controldata_utils.h
@@ -12,7 +12,8 @@
 
 #include "catalog/pg_control.h"
 
-extern ControlFileData *get_controlfile(const char *DataDir, bool *crc_ok_p);
+extern ControlFileData *get_controlfile(const char *ControlFilePath, bool *crc_ok_p);
+extern ControlFileData *get_dir_controlfile(const char *DataDir, bool *crc_ok_p);
 extern void update_controlfile(const char *DataDir,
 							   ControlFileData *ControlFile, bool do_sync);
 
-- 
2.18.0



  [application/x-patch] v8-0004-Add-system-identifier-to-backup-manifest.patch (30.6K, ../../CAAJ_b97_9nJCD3kAmNfSnN1s7m9n1h97OP=APwts=g-rWGztxQ@mail.gmail.com/6-v8-0004-Add-system-identifier-to-backup-manifest.patch)
  download | inline diff:
From da54c83d1a5c11bf978319b3dc6858f2de3e1d4b Mon Sep 17 00:00:00 2001
From: Amul Sul <[email protected]>
Date: Thu, 15 Feb 2024 18:13:29 +0530
Subject: [PATCH v8 4/4] Add system identifier to backup manifest

The backup manifest will be having a new item as "System-Identifier"
and its value is from "ControlFileData.system_identifier" when
manifest generated.  This help identify the correct database server
and/or backup while taking subsequent backup.
---
 doc/src/sgml/backup-manifest.sgml             | 16 +++-
 doc/src/sgml/ref/pg_verifybackup.sgml         |  7 +-
 src/backend/backup/backup_manifest.c          |  7 +-
 src/backend/backup/basebackup_incremental.c   | 39 +++++++++
 src/bin/pg_basebackup/t/010_pg_basebackup.pl  | 16 ++++
 src/bin/pg_combinebackup/load_manifest.c      | 33 +++++++-
 src/bin/pg_combinebackup/load_manifest.h      |  1 +
 src/bin/pg_combinebackup/pg_combinebackup.c   | 34 ++++++--
 src/bin/pg_combinebackup/t/005_integrity.pl   | 14 ++++
 src/bin/pg_combinebackup/write_manifest.c     |  8 +-
 src/bin/pg_combinebackup/write_manifest.h     |  3 +-
 src/bin/pg_verifybackup/pg_verifybackup.c     | 82 ++++++++++++++++++-
 src/bin/pg_verifybackup/t/003_corruption.pl   | 26 +++++-
 src/bin/pg_verifybackup/t/005_bad_manifest.pl | 17 +++-
 src/common/parse_manifest.c                   | 82 ++++++++++++++++++-
 src/include/common/parse_manifest.h           |  6 ++
 16 files changed, 367 insertions(+), 24 deletions(-)

diff --git a/doc/src/sgml/backup-manifest.sgml b/doc/src/sgml/backup-manifest.sgml
index 771be1310a1..c43af36dc6f 100644
--- a/doc/src/sgml/backup-manifest.sgml
+++ b/doc/src/sgml/backup-manifest.sgml
@@ -37,7 +37,21 @@
     <term><literal>PostgreSQL-Backup-Manifest-Version</literal></term>
     <listitem>
      <para>
-      The associated value is always the integer 1.
+      The associated value is an integer. Beginning in
+      <productname>PostgreSQL</productname> <literal>17</literal>,
+      it is <literal>2</literal>; in older versions, it is <literal>1</literal>.
+     </para>
+    </listitem>
+   </varlistentry>
+
+   <varlistentry>
+    <term><literal>System-Identifier</literal></term>
+    <listitem>
+     <para>
+      The associated integer value is an unique system identifier to ensure
+      file match up with the installation that produced them. Available only
+      when <literal>PostgreSQL-Backup-Manifest-Version</literal> is
+      <literal>2</literal> and later.
      </para>
     </listitem>
    </varlistentry>
diff --git a/doc/src/sgml/ref/pg_verifybackup.sgml b/doc/src/sgml/ref/pg_verifybackup.sgml
index 36335e0a188..a3f167f9f6e 100644
--- a/doc/src/sgml/ref/pg_verifybackup.sgml
+++ b/doc/src/sgml/ref/pg_verifybackup.sgml
@@ -53,9 +53,10 @@ PostgreSQL documentation
    Backup verification proceeds in four stages. First,
    <literal>pg_verifybackup</literal> reads the
    <literal>backup_manifest</literal> file. If that file
-   does not exist, cannot be read, is malformed, or fails verification
-   against its own internal checksum, <literal>pg_verifybackup</literal>
-   will terminate with a fatal error.
+   does not exist, cannot be read, is malformed, fails to match the system
+   identifier with <filename>pg_control</filename> of the backup directory or
+   fails verification against its own internal checksum,
+   <literal>pg_verifybackup</literal> will terminate with a fatal error.
   </para>
 
   <para>
diff --git a/src/backend/backup/backup_manifest.c b/src/backend/backup/backup_manifest.c
index 2c34e597523..0f31ae72e78 100644
--- a/src/backend/backup/backup_manifest.c
+++ b/src/backend/backup/backup_manifest.c
@@ -13,6 +13,7 @@
 #include "postgres.h"
 
 #include "access/timeline.h"
+#include "access/xlog.h"
 #include "backup/backup_manifest.h"
 #include "backup/basebackup_sink.h"
 #include "libpq/libpq.h"
@@ -79,8 +80,10 @@ InitializeBackupManifest(backup_manifest_info *manifest,
 
 	if (want_manifest != MANIFEST_OPTION_NO)
 		AppendToManifest(manifest,
-						 "{ \"PostgreSQL-Backup-Manifest-Version\": 1,\n"
-						 "\"Files\": [");
+						 "{ \"PostgreSQL-Backup-Manifest-Version\": 2,\n"
+						 "\"System-Identifier\": " UINT64_FORMAT ",\n"
+						 "\"Files\": [",
+						 GetSystemIdentifier());
 }
 
 /*
diff --git a/src/backend/backup/basebackup_incremental.c b/src/backend/backup/basebackup_incremental.c
index e994ee66bbf..46cf2ad5c74 100644
--- a/src/backend/backup/basebackup_incremental.c
+++ b/src/backend/backup/basebackup_incremental.c
@@ -113,6 +113,10 @@ struct IncrementalBackupInfo
 	BlockRefTable *brtab;
 };
 
+static void manifest_process_version(JsonManifestParseContext *context,
+									 int manifest_version);
+static void manifest_process_system_identifier(JsonManifestParseContext *context,
+											   uint64 manifest_system_identifier);
 static void manifest_process_file(JsonManifestParseContext *context,
 								  char *pathname,
 								  size_t size,
@@ -199,6 +203,8 @@ FinalizeIncrementalManifest(IncrementalBackupInfo *ib)
 
 	/* Parse the manifest. */
 	context.private_data = ib;
+	context.version_cb = manifest_process_version;
+	context.system_identifier_cb = manifest_process_system_identifier;
 	context.per_file_cb = manifest_process_file;
 	context.per_wal_range_cb = manifest_process_wal_range;
 	context.error_cb = manifest_report_error;
@@ -911,6 +917,39 @@ hash_string_pointer(const char *s)
 	return hash_bytes(ss, strlen(s));
 }
 
+/*
+ * This callback to validate the manifest version for incremental backup.
+ */
+static void
+manifest_process_version(JsonManifestParseContext *context,
+						 int manifest_version)
+{
+	/* Incremental backups supported on manifest version 2 or later */
+	if (manifest_version == 1)
+		context->error_cb(context,
+						  "backup manifest version 1 does not support incremental backup");
+}
+
+/*
+ * This callback to validate the manifest system identifier against the current
+ * database server.
+ */
+static void
+manifest_process_system_identifier(JsonManifestParseContext *context,
+								   uint64 manifest_system_identifier)
+{
+	uint64		system_identifier;
+
+	/* Get system identifier of current system */
+	system_identifier = GetSystemIdentifier();
+
+	if (manifest_system_identifier != system_identifier)
+		context->error_cb(context,
+						  "manifest system identifier is %llu, but database system identifier is %llu",
+						  (unsigned long long) manifest_system_identifier,
+						  (unsigned long long) system_identifier);
+}
+
 /*
  * This callback is invoked for each file mentioned in the backup manifest.
  *
diff --git a/src/bin/pg_basebackup/t/010_pg_basebackup.pl b/src/bin/pg_basebackup/t/010_pg_basebackup.pl
index 86cc01a640b..e4592834852 100644
--- a/src/bin/pg_basebackup/t/010_pg_basebackup.pl
+++ b/src/bin/pg_basebackup/t/010_pg_basebackup.pl
@@ -966,4 +966,20 @@ $backupdir = $node->backup_dir . '/backup3';
 my @dst_tblspc = glob "$backupdir/pg_tblspc/$tblspc_oid/PG_*";
 is(@dst_tblspc, 1, 'tblspc directory copied');
 
+# Can't take backup with referring manifest of different cluster
+#
+# Set up another new database instance with force initdb option. We don't want
+# to initializing database system by copying initdb template for this, because
+# we want it to be a separate cluster with a different system ID.
+my $node2 = PostgreSQL::Test::Cluster->new('node2');
+$node2->init(force_initdb => 1, has_archiving => 1, allows_streaming => 1);
+$node2->append_conf('postgresql.conf', 'summarize_wal = on');
+$node2->start;
+
+$node2->command_fails_like(
+	[ @pg_basebackup_defs, '-D', "$tempdir" . '/diff_sysid',
+		'--incremental', "$backupdir" . '/backup_manifest' ],
+	qr/manifest system identifier is .*, but database system identifier is/,
+	"pg_basebackup fails with different database system manifest");
+
 done_testing();
diff --git a/src/bin/pg_combinebackup/load_manifest.c b/src/bin/pg_combinebackup/load_manifest.c
index 2b8e74fcf38..4eae7f61efc 100644
--- a/src/bin/pg_combinebackup/load_manifest.c
+++ b/src/bin/pg_combinebackup/load_manifest.c
@@ -50,6 +50,10 @@ static uint32 hash_string_pointer(char *s);
 #define SH_DEFINE
 #include "lib/simplehash.h"
 
+static void combinebackup_version_cb(JsonManifestParseContext *context,
+									 int manifest_version);
+static void combinebackup_system_identifier_cb(JsonManifestParseContext *context,
+											   uint64 manifest_system_identifier);
 static void combinebackup_per_file_cb(JsonManifestParseContext *context,
 									  char *pathname, size_t size,
 									  pg_checksum_type checksum_type,
@@ -103,8 +107,8 @@ load_backup_manifest(char *backup_directory)
 	manifest_files_hash *ht;
 	char	   *buffer;
 	int			rc;
-	JsonManifestParseContext context;
 	manifest_data *result;
+	JsonManifestParseContext context;
 
 	/* Open the manifest file. */
 	snprintf(pathname, MAXPGPATH, "%s/backup_manifest", backup_directory);
@@ -153,6 +157,8 @@ load_backup_manifest(char *backup_directory)
 	result = pg_malloc0(sizeof(manifest_data));
 	result->files = ht;
 	context.private_data = result;
+	context.version_cb = combinebackup_version_cb;
+	context.system_identifier_cb = combinebackup_system_identifier_cb;
 	context.per_file_cb = combinebackup_per_file_cb;
 	context.per_wal_range_cb = combinebackup_per_wal_range_cb;
 	context.error_cb = report_manifest_error;
@@ -181,6 +187,31 @@ report_manifest_error(JsonManifestParseContext *context, const char *fmt,...)
 	exit(1);
 }
 
+/*
+ * This callback to validate the manifest version number for incremental backup.
+ */
+static void
+combinebackup_version_cb(JsonManifestParseContext *context,
+						 int manifest_version)
+{
+	/* Incremental backups supported on manifest version 2 or later */
+	if (manifest_version == 1)
+		pg_fatal("backup manifest version 1 does not support incremental backup");
+}
+
+/*
+ * Record system identifier extracted from the backup manifest.
+ */
+static void
+combinebackup_system_identifier_cb(JsonManifestParseContext *context,
+								   uint64 manifest_system_identifier)
+{
+	manifest_data *manifest = context->private_data;
+
+	/* Validation will be at the later stage */
+	manifest->system_identifier = manifest_system_identifier;
+}
+
 /*
  * Record details extracted from the backup manifest for one file.
  */
diff --git a/src/bin/pg_combinebackup/load_manifest.h b/src/bin/pg_combinebackup/load_manifest.h
index 9163e071afd..8a5a70e4477 100644
--- a/src/bin/pg_combinebackup/load_manifest.h
+++ b/src/bin/pg_combinebackup/load_manifest.h
@@ -55,6 +55,7 @@ typedef struct manifest_wal_range
  */
 typedef struct manifest_data
 {
+	uint64		system_identifier;
 	manifest_files_hash *files;
 	manifest_wal_range *first_wal_range;
 	manifest_wal_range *last_wal_range;
diff --git a/src/bin/pg_combinebackup/pg_combinebackup.c b/src/bin/pg_combinebackup/pg_combinebackup.c
index 00c018351ac..b5486aad1de 100644
--- a/src/bin/pg_combinebackup/pg_combinebackup.c
+++ b/src/bin/pg_combinebackup/pg_combinebackup.c
@@ -92,7 +92,7 @@ cb_cleanup_dir *cleanup_dir_list = NULL;
 
 static void add_tablespace_mapping(cb_options *opt, char *arg);
 static StringInfo check_backup_label_files(int n_backups, char **backup_dirs);
-static void check_control_files(int n_backups, char **backup_dirs);
+static uint64 check_control_files(int n_backups, char **backup_dirs);
 static void check_input_dir_permissions(char *dir);
 static void cleanup_directories_atexit(void);
 static void create_output_directory(char *dirname, cb_options *opt);
@@ -134,11 +134,13 @@ main(int argc, char *argv[])
 
 	const char *progname;
 	char	   *last_input_dir;
+	int			i;
 	int			optindex;
 	int			c;
 	int			n_backups;
 	int			n_prior_backups;
 	int			version;
+	uint64		system_identifier;
 	char	  **prior_backup_dirs;
 	cb_options	opt;
 	cb_tablespace *tablespaces;
@@ -216,7 +218,7 @@ main(int argc, char *argv[])
 
 	/* Sanity-check control files. */
 	n_backups = argc - optind;
-	check_control_files(n_backups, argv + optind);
+	system_identifier = check_control_files(n_backups, argv + optind);
 
 	/* Sanity-check backup_label files, and get the contents of the last one. */
 	last_backup_label = check_backup_label_files(n_backups, argv + optind);
@@ -231,6 +233,26 @@ main(int argc, char *argv[])
 	/* Load backup manifests. */
 	manifests = load_backup_manifests(n_backups, prior_backup_dirs);
 
+	/*
+	 * Validate the manifest system identifier against the backup system
+	 * identifier.
+	 */
+	for (i = 0; i < n_backups; i++)
+	{
+		if (manifests[i] &&
+			manifests[i]->system_identifier != system_identifier)
+		{
+			char	   *controlpath;
+
+			controlpath = psprintf("%s/%s", prior_backup_dirs[i], "global/pg_control");
+
+			pg_fatal("%s: manifest system identifier is %llu, but control file has %llu",
+					 controlpath,
+					 (unsigned long long) manifests[i]->system_identifier,
+					 (unsigned long long) system_identifier);
+		}
+	}
+
 	/* Figure out which tablespaces are going to be included in the output. */
 	last_input_dir = argv[argc - 1];
 	check_input_dir_permissions(last_input_dir);
@@ -256,7 +278,7 @@ main(int argc, char *argv[])
 	/* If we need to write a backup_manifest, prepare to do so. */
 	if (!opt.dry_run && !opt.no_manifest)
 	{
-		mwriter = create_manifest_writer(opt.output);
+		mwriter = create_manifest_writer(opt.output, system_identifier);
 
 		/*
 		 * Verify that we have a backup manifest for the final backup; else we
@@ -517,9 +539,9 @@ check_backup_label_files(int n_backups, char **backup_dirs)
 }
 
 /*
- * Sanity check control files.
+ * Sanity check control files and return system_identifier.
  */
-static void
+static uint64
 check_control_files(int n_backups, char **backup_dirs)
 {
 	int			i;
@@ -564,6 +586,8 @@ check_control_files(int n_backups, char **backup_dirs)
 	 */
 	pg_log_debug("system identifier is %llu",
 				 (unsigned long long) system_identifier);
+
+	return system_identifier;
 }
 
 /*
diff --git a/src/bin/pg_combinebackup/t/005_integrity.pl b/src/bin/pg_combinebackup/t/005_integrity.pl
index 5d425209211..d79348ab6f3 100644
--- a/src/bin/pg_combinebackup/t/005_integrity.pl
+++ b/src/bin/pg_combinebackup/t/005_integrity.pl
@@ -8,6 +8,7 @@ use strict;
 use warnings FATAL => 'all';
 use File::Compare;
 use File::Path qw(rmtree);
+use File::Copy;
 use PostgreSQL::Test::Cluster;
 use PostgreSQL::Test::Utils;
 use Test::More;
@@ -80,6 +81,19 @@ $node1->command_fails_like(
 	qr/expected system identifier.*but found/,
 	"can't combine backups from different nodes");
 
+# Can't combine when different manifest system identifier
+rename("$backup2path/backup_manifest", "$backup2path/backup_manifest.orig")
+  or BAIL_OUT "could not move $backup2path/backup_manifest";
+copy("$backupother2path/backup_manifest", "$backup2path/backup_manifest")
+  or BAIL_OUT "could not copy $backupother2path/backup_manifest";
+$node1->command_fails_like(
+	[ 'pg_combinebackup', $backup1path, $backup2path, $backup3path, '-o', $resultpath ],
+	qr/ manifest system identifier is .*, but control file has /,
+	"can't combine backups with different manifest system identifier ");
+# Restore the backup state
+move("$backup2path/backup_manifest.orig", "$backup2path/backup_manifest")
+  or BAIL_OUT "could not move $backup2path/backup_manifest";
+
 # Can't omit a required backup.
 $node1->command_fails_like(
 	[ 'pg_combinebackup', $backup1path, $backup3path, '-o', $resultpath ],
diff --git a/src/bin/pg_combinebackup/write_manifest.c b/src/bin/pg_combinebackup/write_manifest.c
index 01deb82cc9f..7a2065e1db7 100644
--- a/src/bin/pg_combinebackup/write_manifest.c
+++ b/src/bin/pg_combinebackup/write_manifest.c
@@ -45,7 +45,7 @@ static size_t hex_encode(const uint8 *src, size_t len, char *dst);
  * in the specified directory.
  */
 manifest_writer *
-create_manifest_writer(char *directory)
+create_manifest_writer(char *directory, uint64 system_identifier)
 {
 	manifest_writer *mwriter = pg_malloc(sizeof(manifest_writer));
 
@@ -57,8 +57,10 @@ create_manifest_writer(char *directory)
 	pg_checksum_init(&mwriter->manifest_ctx, CHECKSUM_TYPE_SHA256);
 
 	appendStringInfo(&mwriter->buf,
-					 "{ \"PostgreSQL-Backup-Manifest-Version\": 1,\n"
-					 "\"Files\": [");
+					 "{ \"PostgreSQL-Backup-Manifest-Version\": 2,\n"
+					 "\"System-Identifier\": " UINT64_FORMAT ",\n"
+					 "\"Files\": [",
+					 system_identifier);
 
 	return mwriter;
 }
diff --git a/src/bin/pg_combinebackup/write_manifest.h b/src/bin/pg_combinebackup/write_manifest.h
index de0f742779f..ebc4f9441ad 100644
--- a/src/bin/pg_combinebackup/write_manifest.h
+++ b/src/bin/pg_combinebackup/write_manifest.h
@@ -19,7 +19,8 @@ struct manifest_wal_range;
 struct manifest_writer;
 typedef struct manifest_writer manifest_writer;
 
-extern manifest_writer *create_manifest_writer(char *directory);
+extern manifest_writer *create_manifest_writer(char *directory,
+											   uint64 system_identifier);
 extern void add_file_to_manifest(manifest_writer *mwriter,
 								 const char *manifest_path,
 								 size_t size, time_t mtime,
diff --git a/src/bin/pg_verifybackup/pg_verifybackup.c b/src/bin/pg_verifybackup/pg_verifybackup.c
index 8561678a7d0..7f730514f3e 100644
--- a/src/bin/pg_verifybackup/pg_verifybackup.c
+++ b/src/bin/pg_verifybackup/pg_verifybackup.c
@@ -18,6 +18,7 @@
 #include <sys/stat.h>
 #include <time.h>
 
+#include "common/controldata_utils.h"
 #include "common/hashfn.h"
 #include "common/logging.h"
 #include "common/parse_manifest.h"
@@ -98,6 +99,8 @@ typedef struct manifest_wal_range
  */
 typedef struct manifest_data
 {
+	int			version;
+	uint64		system_identifier;
 	manifest_files_hash *files;
 	manifest_wal_range *first_wal_range;
 	manifest_wal_range *last_wal_range;
@@ -116,6 +119,10 @@ typedef struct verifier_context
 } verifier_context;
 
 static manifest_data *parse_manifest_file(char *manifest_path);
+static void verifybackup_version_cb(JsonManifestParseContext *context,
+									int manifest_version);
+static void verifybackup_system_identifier(JsonManifestParseContext *context,
+										   uint64 manifest_system_identifier);
 static void verifybackup_per_file_cb(JsonManifestParseContext *context,
 									 char *pathname, size_t size,
 									 pg_checksum_type checksum_type,
@@ -129,6 +136,8 @@ static void report_manifest_error(JsonManifestParseContext *context,
 								  const char *fmt,...)
 			pg_attribute_printf(2, 3) pg_attribute_noreturn();
 
+static void verify_system_identifier(manifest_data *manifest,
+									 char *controlpath);
 static void verify_backup_directory(verifier_context *context,
 									char *relpath, char *fullpath);
 static void verify_backup_file(verifier_context *context,
@@ -375,9 +384,7 @@ main(int argc, char **argv)
 }
 
 /*
- * Parse a manifest file. Construct a hash table with information about
- * all the files it mentions, and a linked list of all the WAL ranges it
- * mentions.
+ * Parse a manifest file and return a data structure describing the contents.
  */
 static manifest_data *
 parse_manifest_file(char *manifest_path)
@@ -432,6 +439,8 @@ parse_manifest_file(char *manifest_path)
 	result = pg_malloc0(sizeof(manifest_data));
 	result->files = ht;
 	context.private_data = result;
+	context.version_cb = verifybackup_version_cb;
+	context.system_identifier_cb = verifybackup_system_identifier;
 	context.per_file_cb = verifybackup_per_file_cb;
 	context.per_wal_range_cb = verifybackup_per_wal_range_cb;
 	context.error_cb = report_manifest_error;
@@ -461,6 +470,32 @@ report_manifest_error(JsonManifestParseContext *context, const char *fmt,...)
 	exit(1);
 }
 
+/*
+ * Record details extracted from the backup manifest.
+ */
+static void
+verifybackup_version_cb(JsonManifestParseContext *context,
+						int manifest_version)
+{
+	manifest_data *manifest = context->private_data;
+
+	/* Validation will be at the later stage */
+	manifest->version = manifest_version;
+}
+
+/*
+ * Record details extracted from the backup manifest.
+ */
+static void
+verifybackup_system_identifier(JsonManifestParseContext *context,
+						 uint64 manifest_system_identifier)
+{
+	manifest_data *manifest = context->private_data;
+
+	/* Validation will be at the later stage */
+	manifest->system_identifier = manifest_system_identifier;
+}
+
 /*
  * Record details extracted from the backup manifest for one file.
  */
@@ -517,6 +552,39 @@ verifybackup_per_wal_range_cb(JsonManifestParseContext *context,
 	manifest->last_wal_range = range;
 }
 
+/*
+ * Sanity check control file of the backup and validate system identifier
+ * against manifest system identifier.
+ */
+static void
+verify_system_identifier(manifest_data *manifest, char *controlpath)
+{
+	ControlFileData *control_file;
+	bool		crc_ok;
+
+	pg_log_debug("reading \"%s\"", controlpath);
+	control_file = get_controlfile(controlpath, &crc_ok);
+
+	/* Control file contents not meaningful if CRC is bad. */
+	if (!crc_ok)
+		report_fatal_error("%s: CRC is incorrect", controlpath);
+
+	/* Can't interpret control file if not current version. */
+	if (control_file->pg_control_version != PG_CONTROL_VERSION)
+		report_fatal_error("%s: unexpected control file version",
+						   controlpath);
+
+	/* System identifiers should match. */
+	if (manifest->system_identifier != control_file->system_identifier)
+		report_fatal_error("%s: manifest system identifier is %llu, but control file has %llu",
+						   controlpath,
+						   (unsigned long long) manifest->system_identifier,
+						   (unsigned long long) control_file->system_identifier);
+
+	/* Release memory. */
+	pfree(control_file);
+}
+
 /*
  * Verify one directory.
  *
@@ -650,6 +718,14 @@ verify_backup_file(verifier_context *context, char *relpath, char *fullpath)
 		m->bad = true;
 	}
 
+	/*
+	 * Validate the manifest system identifier, not available in manifest
+	 * version 1.
+	 */
+	if (context->manifest->version != 1 &&
+		strcmp(relpath, "global/pg_control") == 0)
+			verify_system_identifier(context->manifest, fullpath);
+
 	/* Update statistics for progress report, if necessary */
 	if (show_progress && !skip_checksums && should_verify_checksum(m))
 		total_size += m->size;
diff --git a/src/bin/pg_verifybackup/t/003_corruption.pl b/src/bin/pg_verifybackup/t/003_corruption.pl
index 11bd5770818..36d032288fe 100644
--- a/src/bin/pg_verifybackup/t/003_corruption.pl
+++ b/src/bin/pg_verifybackup/t/003_corruption.pl
@@ -6,6 +6,7 @@
 use strict;
 use warnings FATAL => 'all';
 use File::Path qw(rmtree);
+use File::Copy;
 use PostgreSQL::Test::Cluster;
 use PostgreSQL::Test::Utils;
 use Test::More;
@@ -68,6 +69,11 @@ my @scenario = (
 		'mutilate' => \&mutilate_replace_file,
 		'fails_like' => qr/checksum mismatch for file/
 	},
+	{
+		'name' => 'system_identifier',
+		'mutilate' => \&mutilate_system_identifier,
+		'fails_like' => qr/manifest system identifier is .*, but control file has/
+	},
 	{
 		'name' => 'bad_manifest',
 		'mutilate' => \&mutilate_bad_manifest,
@@ -216,7 +222,7 @@ sub mutilate_append_to_file
 sub mutilate_truncate_file
 {
 	my ($backup_path) = @_;
-	my $pathname = "$backup_path/global/pg_control";
+	my $pathname = "$backup_path/pg_hba.conf";
 	open(my $fh, '>', $pathname) || die "open $pathname: $!";
 	close($fh);
 	return;
@@ -236,6 +242,24 @@ sub mutilate_replace_file
 	return;
 }
 
+# Copy manifest of other backups to demonstrate the case where the wrong
+# manifest is referred
+sub mutilate_system_identifier
+{
+	my ($backup_path) = @_;
+
+	# Set up another new database instance with different system identifier and
+	# make backup
+	my $node = PostgreSQL::Test::Cluster->new('node');
+	$node->init(force_initdb => 1, allows_streaming => 1);
+	$node->start;
+	$node->backup('backup2');
+	move($node->backup_dir.'/backup2/backup_manifest', $backup_path.'/backup_manifest')
+		or BAIL_OUT "could not copy manifest to $backup_path";
+	$node->teardown_node(fail_ok => 1);
+	return;
+}
+
 # Corrupt the backup manifest.
 sub mutilate_bad_manifest
 {
diff --git a/src/bin/pg_verifybackup/t/005_bad_manifest.pl b/src/bin/pg_verifybackup/t/005_bad_manifest.pl
index e278ccea5a2..f95b45a7230 100644
--- a/src/bin/pg_verifybackup/t/005_bad_manifest.pl
+++ b/src/bin/pg_verifybackup/t/005_bad_manifest.pl
@@ -29,10 +29,14 @@ test_parse_error('expected version indicator', <<EOM);
 {"not-expected": 1}
 EOM
 
-test_parse_error('unexpected manifest version', <<EOM);
+test_parse_error('manifest version not an integer', <<EOM);
 {"PostgreSQL-Backup-Manifest-Version": "phooey"}
 EOM
 
+test_parse_error('unexpected manifest version', <<EOM);
+{"PostgreSQL-Backup-Manifest-Version": 9876599}
+EOM
+
 test_parse_error('unexpected scalar', <<EOM);
 {"PostgreSQL-Backup-Manifest-Version": 1, "Files": true}
 EOM
@@ -156,6 +160,17 @@ test_parse_error('expected at least 2 lines', <<EOM);
 {"PostgreSQL-Backup-Manifest-Version": 1, "Files": [], "Manifest-Checksum": null}
 EOM
 
+test_parse_error('unrecognized top-level field', <<EOM);
+{"PostgreSQL-Backup-Manifest-Version": 1,
+ "System-Identifier": "7320280665284567118"
+ "Manifest-Checksum": "eabdc9caf8a" }
+EOM
+
+test_parse_error('expected system identifier', <<EOM);
+{"PostgreSQL-Backup-Manifest-Version": 2,
+ "Manifest-Checksum": "eabdc9caf8a" }
+EOM
+
 my $manifest_without_newline = <<EOM;
 {"PostgreSQL-Backup-Manifest-Version": 1,
  "Files": [],
diff --git a/src/common/parse_manifest.c b/src/common/parse_manifest.c
index 92a97714f36..7e3682f75b0 100644
--- a/src/common/parse_manifest.c
+++ b/src/common/parse_manifest.c
@@ -25,6 +25,7 @@ typedef enum
 	JM_EXPECT_TOPLEVEL_END,
 	JM_EXPECT_TOPLEVEL_FIELD,
 	JM_EXPECT_VERSION_VALUE,
+	JM_EXPECT_SYSTEM_IDENTIFIER_VALUE,
 	JM_EXPECT_FILES_START,
 	JM_EXPECT_FILES_NEXT,
 	JM_EXPECT_THIS_FILE_FIELD,
@@ -85,6 +86,9 @@ typedef struct
 
 	/* Miscellaneous other stuff. */
 	bool		saw_version_field;
+	bool		saw_system_identifier_field;
+	char	   *manifest_version;
+	char	   *manifest_system_identifier;
 	char	   *manifest_checksum;
 } JsonManifestParseState;
 
@@ -96,6 +100,8 @@ static JsonParseErrorType json_manifest_object_field_start(void *state, char *fn
 														   bool isnull);
 static JsonParseErrorType json_manifest_scalar(void *state, char *token,
 											   JsonTokenType tokentype);
+static void json_manifest_finalize_version(JsonManifestParseState *parse);
+static void json_manifest_finalize_system_identifier(JsonManifestParseState *parse);
 static void json_manifest_finalize_file(JsonManifestParseState *parse);
 static void json_manifest_finalize_wal_range(JsonManifestParseState *parse);
 static void verify_manifest_checksum(JsonManifestParseState *parse,
@@ -128,6 +134,7 @@ json_parse_manifest(JsonManifestParseContext *context, char *buffer,
 	parse.context = context;
 	parse.state = JM_EXPECT_TOPLEVEL_START;
 	parse.saw_version_field = false;
+	parse.saw_system_identifier_field = false;
 
 	/* Create a JSON lexing context. */
 	lex = makeJsonLexContextCstringLen(NULL, buffer, size, PG_UTF8, true);
@@ -311,6 +318,16 @@ json_manifest_object_field_start(void *state, char *fname, bool isnull)
 				parse->saw_version_field = true;
 				break;
 			}
+			else if (!parse->saw_system_identifier_field &&
+					 strcmp(parse->manifest_version, "1") != 0)
+			{
+				if (strcmp(fname, "System-Identifier") != 0)
+					json_manifest_parse_failure(parse->context,
+												"expected system identifier");
+				parse->state = JM_EXPECT_SYSTEM_IDENTIFIER_VALUE;
+				parse->saw_system_identifier_field = true;
+				break;
+			}
 
 			/* Is this the list of files? */
 			if (strcmp(fname, "Files") == 0)
@@ -404,9 +421,14 @@ json_manifest_scalar(void *state, char *token, JsonTokenType tokentype)
 	switch (parse->state)
 	{
 		case JM_EXPECT_VERSION_VALUE:
-			if (strcmp(token, "1") != 0)
-				json_manifest_parse_failure(parse->context,
-											"unexpected manifest version");
+			parse->manifest_version = token;
+			json_manifest_finalize_version(parse);
+			parse->state = JM_EXPECT_TOPLEVEL_FIELD;
+			break;
+
+		case JM_EXPECT_SYSTEM_IDENTIFIER_VALUE:
+			parse->manifest_system_identifier = token;
+			json_manifest_finalize_system_identifier(parse);
 			parse->state = JM_EXPECT_TOPLEVEL_FIELD;
 			break;
 
@@ -464,6 +486,60 @@ json_manifest_scalar(void *state, char *token, JsonTokenType tokentype)
 	return JSON_SUCCESS;
 }
 
+/*
+ * Do additional parsing and sanity-checking of the manifest version, and invoke
+ * the callback so that the caller can gets that detail and take actions
+ * accordingly.  This happens for each manifest when the corresponding JSON
+ * object is completely parsed.
+ */
+static void
+json_manifest_finalize_version(JsonManifestParseState *parse)
+{
+	JsonManifestParseContext *context = parse->context;
+	int			version;
+	char	   *ep;
+
+	Assert(parse->saw_version_field);
+
+	/* Parse version. */
+	version = strtoi64(parse->manifest_version, &ep, 10);
+	if (*ep)
+		json_manifest_parse_failure(parse->context,
+									"manifest version not an integer");
+
+	if (version != 1 && version != 2)
+		json_manifest_parse_failure(parse->context,
+									"unexpected manifest version");
+
+	/* Invoke the callback for version */
+	context->version_cb(context, version);
+}
+
+/*
+ * Do additional parsing and sanity-checking of the system identifier, and
+ * invoke the callback so that the caller can gets that detail and take actions
+ * accordingly.  This happens for each manifest when the corresponding JSON
+ * object is completely parsed.
+ */
+static void
+json_manifest_finalize_system_identifier(JsonManifestParseState *parse)
+{
+	JsonManifestParseContext *context = parse->context;
+	uint64		system_identifier;
+	char	   *ep;
+
+	Assert(parse->saw_system_identifier_field);
+
+	/* Parse system identifier. */
+	system_identifier = strtou64(parse->manifest_system_identifier, &ep, 10);
+	if (*ep)
+		json_manifest_parse_failure(parse->context,
+									"manifest system identifier not an integer");
+
+	/* Invoke the callback for system identifier */
+	context->system_identifier_cb(context, system_identifier);
+}
+
 /*
  * Do additional parsing and sanity-checking of the details gathered for one
  * file, and invoke the per-file callback so that the caller gets those
diff --git a/src/include/common/parse_manifest.h b/src/include/common/parse_manifest.h
index f74be0db35f..78b052c045b 100644
--- a/src/include/common/parse_manifest.h
+++ b/src/include/common/parse_manifest.h
@@ -21,6 +21,10 @@
 struct JsonManifestParseContext;
 typedef struct JsonManifestParseContext JsonManifestParseContext;
 
+typedef void (*json_manifest_version_callback) (JsonManifestParseContext *,
+												int manifest_version);
+typedef void (*json_manifest_system_identifier_callback) (JsonManifestParseContext *,
+														  uint64 manifest_system_identifier);
 typedef void (*json_manifest_per_file_callback) (JsonManifestParseContext *,
 												 char *pathname,
 												 size_t size, pg_checksum_type checksum_type,
@@ -35,6 +39,8 @@ typedef void (*json_manifest_error_callback) (JsonManifestParseContext *,
 struct JsonManifestParseContext
 {
 	void	   *private_data;
+	json_manifest_version_callback version_cb;
+	json_manifest_system_identifier_callback system_identifier_cb;
 	json_manifest_per_file_callback per_file_cb;
 	json_manifest_per_wal_range_callback per_wal_range_cb;
 	json_manifest_error_callback error_cb;
-- 
2.18.0



^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* Re: Add system identifier to backup manifest
  2024-02-15 09:35 Re: Add system identifier to backup manifest Amul Sul <[email protected]>
  2024-02-15 12:11 ` Re: Add system identifier to backup manifest Robert Haas <[email protected]>
  2024-02-18 22:52   ` Re: Add system identifier to backup manifest Michael Paquier <[email protected]>
  2024-02-19 06:36     ` Re: Add system identifier to backup manifest Amul Sul <[email protected]>
@ 2024-02-21 04:31       ` Michael Paquier <[email protected]>
  1 sibling, 0 replies; 281+ messages in thread

From: Michael Paquier @ 2024-02-21 04:31 UTC (permalink / raw)
  To: Amul Sul <[email protected]>; +Cc: Robert Haas <[email protected]>; pgsql-hackers

On Mon, Feb 19, 2024 at 12:06:19PM +0530, Amul Sul wrote:
> On Mon, Feb 19, 2024 at 4:22 AM Michael Paquier <[email protected]> wrote:
>> And the new option should be documented at the top of the init()
>> routine for perldoc.
> 
> Added in the attached version.

I've done some wordsmithing on 0001 and it is OK, so I've applied it
to move the needle.  Hope that helps.
--
Michael


Attachments:

  [application/pgp-signature] signature.asc (833B, ../../[email protected]/2-signature.asc)
  download

^ permalink  raw  reply  [nested|flat] 281+ messages in thread

* Re: Add system identifier to backup manifest
  2024-02-15 09:35 Re: Add system identifier to backup manifest Amul Sul <[email protected]>
  2024-02-15 12:11 ` Re: Add system identifier to backup manifest Robert Haas <[email protected]>
  2024-02-18 22:52   ` Re: Add system identifier to backup manifest Michael Paquier <[email protected]>
  2024-02-19 06:36     ` Re: Add system identifier to backup manifest Amul Sul <[email protected]>
@ 2024-03-01 05:58       ` Michael Paquier <[email protected]>
  1 sibling, 0 replies; 281+ messages in thread

From: Michael Paquier @ 2024-03-01 05:58 UTC (permalink / raw)
  To: Amul Sul <[email protected]>; +Cc: Robert Haas <[email protected]>; pgsql-hackers

On Mon, Feb 19, 2024 at 12:06:19PM +0530, Amul Sul wrote:
> Agreed, now they will have an error as _could not read file "<DataDir>": Is
> a directory_. But, IIUC, that what usually happens with the dev version, and
> the extension needs to be updated for compatibility with the newer version for
> the same reason.

I was reading through the remaining pieces of the patch set, and are
you sure that there is a need for 0002 at all?  The only reason why
get_dir_controlfile() is introduced is to be able to get the contents
of a control file with a full path to it, and not a data folder.  Now,
if I look closely, with 0002~0004 applied, the only two callers of
get_controlfile() are pg_combinebackup.c and pg_verifybackup.c.  Both
of them have an access to the backup directories, which point to the
root of the data folder.  pg_combinebackup can continue to use
backup_dirs[i].  pg_verifybackup has an access to the backup directory
in the context data, if I'm reading this right, so you could just use
that in verify_system_identifier().
--
Michael


Attachments:

  [application/pgp-signature] signature.asc (833B, ../../[email protected]/2-signature.asc)
  download

^ permalink  raw  reply  [nested|flat] 281+ messages in thread


end of thread, other threads:[~2024-03-01 05:58 UTC | newest]

Thread overview: 281+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2018-11-07 07:53 [PATCH v24 3/5] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2018-11-07 07:53 [PATCH 3/5] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v31 4/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v43 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v40 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v32 4/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v42 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v25 4/8] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v38 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v34 4/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v35 4/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v37 3/6] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v41 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v39 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v30 4/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v33 4/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v33 4/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v29 4/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v44 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v52 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v36 4/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v47 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v46 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v45 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v49 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v48 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v51 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2020-03-13 07:59 [PATCH v50 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v55 3/3] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v53 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2021-03-11 09:01 [PATCH v54 3/7] Make archiver process an auxiliary process Kyotaro Horiguchi <[email protected]>
2024-02-15 09:35 Re: Add system identifier to backup manifest Amul Sul <[email protected]>
2024-02-15 12:11 ` Re: Add system identifier to backup manifest Robert Haas <[email protected]>
2024-02-18 22:52   ` Re: Add system identifier to backup manifest Michael Paquier <[email protected]>
2024-02-19 06:36     ` Re: Add system identifier to backup manifest Amul Sul <[email protected]>
2024-02-21 04:31       ` Re: Add system identifier to backup manifest Michael Paquier <[email protected]>
2024-03-01 05:58       ` Re: Add system identifier to backup manifest Michael Paquier <[email protected]>

This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox