agora inbox for pgsql-hackers@postgresql.org  
help / color / mirror / Atom feed
[PATCH v18 1/4] Introduce custodian.
24+ messages / 2 participants
[nested] [flat]

* [PATCH v7 1/6] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 383 ++++++++++++++++++++++++
 src/backend/postmaster/postmaster.c     |  44 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 12 files changed, 488 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 39ac4490db..620a0b1bae 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..e90f5d0d1f
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,383 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(bool retry);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks(true);
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If retry is true, the custodian will re-enqueue the currently running task if
+ * an exception is encountered.
+ */
+static void
+DoCustodianTasks(bool retry)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (retry)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * If immediate is true, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  This is
+ * mostly useful for single-user mode.  If immediate is false, the task is added
+ * to the custodian's queue if it is not already enqueued, and this function
+ * returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, bool immediate, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (immediate)
+		DoCustodianTasks(false);
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index d7257e4056..1f707d64ac 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -251,6 +251,7 @@ bool		remove_temp_files_after_crash = true;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1822,13 +1824,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2768,6 +2773,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3088,6 +3095,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3181,6 +3190,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3638,6 +3661,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(CheckpointerPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) CustodianPID)));
+		signal_child(CustodianPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3815,6 +3850,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3852,6 +3890,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3941,6 +3980,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4134,6 +4174,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 1a6f527051..b19d743cab 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -129,6 +130,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -277,6 +279,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 37aaab1338..f297f489c9 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index 87c15b9c6f..469768c4e4 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb43b2c5e5..1210c2e7a3 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -273,6 +273,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_STARTUP:
 			backendDesc = "startup";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 0af130fbc5..ffe9404c68 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -330,6 +330,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_STARTUP,
 	B_WAL_RECEIVER,
 	B_WAL_SENDER,
@@ -433,6 +434,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -445,6 +447,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..170ca61a21
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, bool immediate, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 2579e619eb..467421e371 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -394,6 +394,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -411,11 +413,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index b578e2ec75..7524e197e5 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--/04w6evG8XlLl3ft
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v7-0002-Also-remove-pgsql_tmp-directories-during-startup.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v18 1/4] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 doc/src/sgml/glossary.sgml              |  11 +
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 377 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 14 files changed, 488 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/doc/src/sgml/glossary.sgml b/doc/src/sgml/glossary.sgml
index 7c01a541fe..ad3f53e2a3 100644
--- a/doc/src/sgml/glossary.sgml
+++ b/doc/src/sgml/glossary.sgml
@@ -144,6 +144,7 @@
      (but not the autovacuum workers),
      the <glossterm linkend="glossary-background-writer">background writer</glossterm>,
      the <glossterm linkend="glossary-checkpointer">checkpointer</glossterm>,
+     the <glossterm linkend="glossary-custodian">custodian</glossterm>,
      the <glossterm linkend="glossary-logger">logger</glossterm>,
      the <glossterm linkend="glossary-startup-process">startup process</glossterm>,
      the <glossterm linkend="glossary-wal-archiver">WAL archiver</glossterm>,
@@ -484,6 +485,16 @@
    </glossdef>
   </glossentry>
 
+  <glossentry id="glossary-custodian">
+   <glossterm>Custodian (process)</glossterm>
+   <glossdef>
+    <para>
+     An <glossterm linkend="glossary-auxiliary-proc">auxiliary process</glossterm>
+     that is responsible for executing assorted cleanup tasks.
+    </para>
+   </glossdef>
+  </glossentry>
+
   <glossentry>
    <glossterm>Data area</glossterm>
    <glosssee otherterm="glossary-data-directory" />
diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..e5af958999
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,377 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(void);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Files(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks();
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If we are not in a standalone backend, the custodian will re-enqueue the
+ * currently running task if an exception is encountered.
+ */
+static void
+DoCustodianTasks(void)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (IsPostmasterEnvironment)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * In standalone backends, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  Otherwise,
+ * the task is added to the custodian's queue if it is not already enqueued,
+ * and this function returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (!IsPostmasterEnvironment)
+		DoCustodianTasks();
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index a8a246921f..6a74423172 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -240,6 +240,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -537,6 +538,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1808,13 +1810,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2728,6 +2733,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3025,6 +3032,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3118,6 +3127,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3532,6 +3555,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3685,6 +3714,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3722,6 +3754,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3815,6 +3848,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4027,6 +4061,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index b1c35653fc..6a8485e865 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index b2abd75ddb..63fd242b1e 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..73d0bc5f02
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index aa13e1d66e..8f0e696663 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 0b2100be4a..48602c8a16 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--LZvS9be/3tNcYl/X
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v18-0002-Move-removal-of-old-serialized-snapshots-to-cust.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v14 1/3] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 382 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 482 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..a94381bc21
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,382 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(void);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks();
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If we are not in a standalone backend, the custodian will re-enqueue the
+ * currently running task if an exception is encountered.
+ */
+static void
+DoCustodianTasks(void)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (IsPostmasterEnvironment)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * In standalone backends, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  Otherwise,
+ * the task is added to the custodian's queue if it is not already enqueued,
+ * and this function returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (!IsPostmasterEnvironment)
+		DoCustodianTasks();
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index a8a246921f..6a74423172 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -240,6 +240,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -537,6 +538,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1808,13 +1810,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2728,6 +2733,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3025,6 +3032,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3118,6 +3127,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3532,6 +3555,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3685,6 +3714,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3722,6 +3754,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3815,6 +3848,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4027,6 +4061,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index b1c35653fc..6a8485e865 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index b2abd75ddb..63fd242b1e 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..73d0bc5f02
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index aa13e1d66e..8f0e696663 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 0b2100be4a..48602c8a16 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--BXVAT5kNtrzKuDFl
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v14-0002-Move-removal-of-old-serialized-snapshots-to-cust.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v12 1/6] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 383 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  44 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 489 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..e90f5d0d1f
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,383 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(bool retry);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks(true);
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If retry is true, the custodian will re-enqueue the currently running task if
+ * an exception is encountered.
+ */
+static void
+DoCustodianTasks(bool retry)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (retry)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * If immediate is true, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  This is
+ * mostly useful for single-user mode.  If immediate is false, the task is added
+ * to the custodian's queue if it is not already enqueued, and this function
+ * returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, bool immediate, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (immediate)
+		DoCustodianTasks(false);
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 0b637ba6a2..3706eec25e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -248,6 +248,7 @@ bool		remove_temp_files_after_crash = true;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -544,6 +545,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1821,13 +1823,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2746,6 +2751,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3066,6 +3073,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3159,6 +3168,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3616,6 +3639,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(CheckpointerPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) CustodianPID)));
+		signal_child(CustodianPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3793,6 +3828,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3830,6 +3868,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3919,6 +3958,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4113,6 +4153,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 13fa07b0ff..1bae34d1ee 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index 92f24a6c9b..d8e6ea45bc 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..170ca61a21
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, bool immediate, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 8d096fdeeb..448dde0161 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 6f2d5612e0..58455dc016 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--UlVJffcvxoiEqYs2
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v12-0002-Also-remove-pgsql_tmp-directories-during-startup.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v13 1/6] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 383 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 483 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..e90f5d0d1f
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,383 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(bool retry);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks(true);
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If retry is true, the custodian will re-enqueue the currently running task if
+ * an exception is encountered.
+ */
+static void
+DoCustodianTasks(bool retry)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (retry)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * If immediate is true, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  This is
+ * mostly useful for single-user mode.  If immediate is false, the task is added
+ * to the custodian's queue if it is not already enqueued, and this function
+ * returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, bool immediate, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (immediate)
+		DoCustodianTasks(false);
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index c83cc8cc6c..00d18ee761 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -240,6 +240,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -537,6 +538,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1808,13 +1810,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2728,6 +2733,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3025,6 +3032,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3118,6 +3127,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3532,6 +3555,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3685,6 +3714,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3722,6 +3754,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3815,6 +3848,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4027,6 +4061,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index b1c35653fc..6a8485e865 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index b2abd75ddb..63fd242b1e 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..170ca61a21
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, bool immediate, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index aa13e1d66e..8f0e696663 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 0b2100be4a..48602c8a16 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--CE+1k2dSO48ffgeK
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v13-0002-Also-remove-pgsql_tmp-directories-during-startup.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v16 1/4] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 382 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 482 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..a94381bc21
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,382 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(void);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks();
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If we are not in a standalone backend, the custodian will re-enqueue the
+ * currently running task if an exception is encountered.
+ */
+static void
+DoCustodianTasks(void)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (IsPostmasterEnvironment)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * In standalone backends, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  Otherwise,
+ * the task is added to the custodian's queue if it is not already enqueued,
+ * and this function returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (!IsPostmasterEnvironment)
+		DoCustodianTasks();
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index a8a246921f..6a74423172 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -240,6 +240,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -537,6 +538,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1808,13 +1810,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2728,6 +2733,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3025,6 +3032,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3118,6 +3127,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3532,6 +3555,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3685,6 +3714,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3722,6 +3754,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3815,6 +3848,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4027,6 +4061,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index b1c35653fc..6a8485e865 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index b2abd75ddb..63fd242b1e 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..73d0bc5f02
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index aa13e1d66e..8f0e696663 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 0b2100be4a..48602c8a16 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--lrZ03NoBR/3+SXJZ
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v16-0002-Move-removal-of-old-serialized-snapshots-to-cust.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v15 1/4] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 382 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 482 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..a94381bc21
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,382 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(void);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks();
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If we are not in a standalone backend, the custodian will re-enqueue the
+ * currently running task if an exception is encountered.
+ */
+static void
+DoCustodianTasks(void)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (IsPostmasterEnvironment)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * In standalone backends, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  Otherwise,
+ * the task is added to the custodian's queue if it is not already enqueued,
+ * and this function returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (!IsPostmasterEnvironment)
+		DoCustodianTasks();
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index a8a246921f..6a74423172 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -240,6 +240,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -537,6 +538,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1808,13 +1810,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2728,6 +2733,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3025,6 +3032,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3118,6 +3127,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3532,6 +3555,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3685,6 +3714,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3722,6 +3754,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3815,6 +3848,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4027,6 +4061,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index b1c35653fc..6a8485e865 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index b2abd75ddb..63fd242b1e 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..73d0bc5f02
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index aa13e1d66e..8f0e696663 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 0b2100be4a..48602c8a16 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--HcAYCG3uE/tztfnV
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v15-0002-Move-removal-of-old-serialized-snapshots-to-cust.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v17 1/4] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 382 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 482 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..a94381bc21
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,382 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(void);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks();
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If we are not in a standalone backend, the custodian will re-enqueue the
+ * currently running task if an exception is encountered.
+ */
+static void
+DoCustodianTasks(void)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (IsPostmasterEnvironment)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * In standalone backends, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  Otherwise,
+ * the task is added to the custodian's queue if it is not already enqueued,
+ * and this function returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (!IsPostmasterEnvironment)
+		DoCustodianTasks();
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index a8a246921f..6a74423172 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -240,6 +240,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -537,6 +538,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1808,13 +1810,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2728,6 +2733,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3025,6 +3032,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3118,6 +3127,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3532,6 +3555,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3685,6 +3714,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3722,6 +3754,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3815,6 +3848,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4027,6 +4061,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index b1c35653fc..6a8485e865 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index b2abd75ddb..63fd242b1e 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..73d0bc5f02
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index aa13e1d66e..8f0e696663 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 0b2100be4a..48602c8a16 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--ew6BAiZeqk4r7MaW
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v17-0002-Move-removal-of-old-serialized-snapshots-to-cust.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v19 1/4] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 doc/src/sgml/glossary.sgml              |  11 +
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 377 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 14 files changed, 488 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/doc/src/sgml/glossary.sgml b/doc/src/sgml/glossary.sgml
index 7c01a541fe..ad3f53e2a3 100644
--- a/doc/src/sgml/glossary.sgml
+++ b/doc/src/sgml/glossary.sgml
@@ -144,6 +144,7 @@
      (but not the autovacuum workers),
      the <glossterm linkend="glossary-background-writer">background writer</glossterm>,
      the <glossterm linkend="glossary-checkpointer">checkpointer</glossterm>,
+     the <glossterm linkend="glossary-custodian">custodian</glossterm>,
      the <glossterm linkend="glossary-logger">logger</glossterm>,
      the <glossterm linkend="glossary-startup-process">startup process</glossterm>,
      the <glossterm linkend="glossary-wal-archiver">WAL archiver</glossterm>,
@@ -484,6 +485,16 @@
    </glossdef>
   </glossentry>
 
+  <glossentry id="glossary-custodian">
+   <glossterm>Custodian (process)</glossterm>
+   <glossdef>
+    <para>
+     An <glossterm linkend="glossary-auxiliary-proc">auxiliary process</glossterm>
+     that is responsible for executing assorted cleanup tasks.
+    </para>
+   </glossdef>
+  </glossentry>
+
   <glossentry>
    <glossterm>Data area</glossterm>
    <glosssee otherterm="glossary-data-directory" />
diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index cae6feb356..a1f042f13a 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..98bb9efcfd
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,377 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(void);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Files(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	sigprocmask(SIG_SETMASK, &UnBlockSig, NULL);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks();
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If we are not in a standalone backend, the custodian will re-enqueue the
+ * currently running task if an exception is encountered.
+ */
+static void
+DoCustodianTasks(void)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (IsPostmasterEnvironment)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * In standalone backends, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  Otherwise,
+ * the task is added to the custodian's queue if it is not already enqueued,
+ * and this function returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (!IsPostmasterEnvironment)
+		DoCustodianTasks();
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 9079922de7..63f2abe3a1 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -6,6 +6,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 2552327d90..e3aef4081e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -249,6 +249,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -560,6 +561,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1795,13 +1797,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2732,6 +2737,8 @@ process_pm_reload_request(void)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3085,6 +3092,8 @@ process_pm_child_exit(void)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3178,6 +3187,20 @@ process_pm_child_exit(void)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3590,6 +3613,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3744,6 +3773,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3781,6 +3813,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3877,6 +3910,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4092,6 +4126,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8f1ded7338..4268a941ad 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 22b4278610..40a83636fe 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -178,6 +178,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index 6e4599278c..9348b441ba 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index 59532bbd80..25c2ba97b8 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -283,6 +283,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 96b3a1e1a0..738e7f8fb0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -324,6 +324,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -430,6 +431,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -442,6 +444,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..73d0bc5f02
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 4258cd92c9..25e00a14ff 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 6cacd6edaf..ea8ba623e9 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_LOGICAL_PARALLEL_APPLY_MAIN,
-- 
2.25.1


--x+6KMIRAuhnl3hBn
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v19-0002-Move-removal-of-old-serialized-snapshots-to-cust.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v20 1/4] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 doc/src/sgml/glossary.sgml              |  11 +
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 377 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/pgstat_io.c  |   4 +-
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 15 files changed, 491 insertions(+), 6 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/doc/src/sgml/glossary.sgml b/doc/src/sgml/glossary.sgml
index 7c01a541fe..ad3f53e2a3 100644
--- a/doc/src/sgml/glossary.sgml
+++ b/doc/src/sgml/glossary.sgml
@@ -144,6 +144,7 @@
      (but not the autovacuum workers),
      the <glossterm linkend="glossary-background-writer">background writer</glossterm>,
      the <glossterm linkend="glossary-checkpointer">checkpointer</glossterm>,
+     the <glossterm linkend="glossary-custodian">custodian</glossterm>,
      the <glossterm linkend="glossary-logger">logger</glossterm>,
      the <glossterm linkend="glossary-startup-process">startup process</glossterm>,
      the <glossterm linkend="glossary-wal-archiver">WAL archiver</glossterm>,
@@ -484,6 +485,16 @@
    </glossdef>
   </glossentry>
 
+  <glossentry id="glossary-custodian">
+   <glossterm>Custodian (process)</glossterm>
+   <glossdef>
+    <para>
+     An <glossterm linkend="glossary-auxiliary-proc">auxiliary process</glossterm>
+     that is responsible for executing assorted cleanup tasks.
+    </para>
+   </glossdef>
+  </glossentry>
+
   <glossentry>
    <glossterm>Data area</glossterm>
    <glosssee otherterm="glossary-data-directory" />
diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 047448b34e..5f4dde85cf 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index cae6feb356..a1f042f13a 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..98bb9efcfd
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,377 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(void);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Files(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	sigprocmask(SIG_SETMASK, &UnBlockSig, NULL);
+
+	/*
+	 * Advertise our latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks();
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If we are not in a standalone backend, the custodian will re-enqueue the
+ * currently running task if an exception is encountered.
+ */
+static void
+DoCustodianTasks(void)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (IsPostmasterEnvironment)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * In standalone backends, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  Otherwise,
+ * the task is added to the custodian's queue if it is not already enqueued,
+ * and this function returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (!IsPostmasterEnvironment)
+		DoCustodianTasks();
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index cda921fd10..faaaba6a21 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -6,6 +6,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 2552327d90..e3aef4081e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -249,6 +249,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -560,6 +561,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1795,13 +1797,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2732,6 +2737,8 @@ process_pm_reload_request(void)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3085,6 +3092,8 @@ process_pm_child_exit(void)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3178,6 +3187,20 @@ process_pm_child_exit(void)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3590,6 +3613,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3744,6 +3773,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3781,6 +3813,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3877,6 +3910,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4092,6 +4126,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8f1ded7338..4268a941ad 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 22b4278610..40a83636fe 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -178,6 +178,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/pgstat_io.c b/src/backend/utils/activity/pgstat_io.c
index 0e07e0848d..f05590f2c4 100644
--- a/src/backend/utils/activity/pgstat_io.c
+++ b/src/backend/utils/activity/pgstat_io.c
@@ -224,7 +224,8 @@ pgstat_io_snapshot_cb(void)
 * - Syslogger because it is not connected to shared memory
 * - Archiver because most relevant archiving IO is delegated to a
 *   specialized command or module
-* - WAL Receiver and WAL Writer IO is not tracked in pg_stat_io for now
+* - WAL Receiver, WAL Writer, and Custodian IO is not tracked in pg_stat_io for
+*   now
 *
 * Function returns true if BackendType participates in the cumulative stats
 * subsystem for IO and false if it does not.
@@ -243,6 +244,7 @@ pgstat_tracks_io_bktype(BackendType bktype)
 	{
 		case B_INVALID:
 		case B_ARCHIVER:
+		case B_CUSTODIAN:
 		case B_LOGGER:
 		case B_WAL_RECEIVER:
 		case B_WAL_WRITER:
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index cb99cc6339..6ca751dd1f 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index 59532bbd80..25c2ba97b8 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -283,6 +283,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index c309e0233d..50f407a0bf 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -324,6 +324,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -432,6 +433,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -444,6 +446,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..73d0bc5f02
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 4258cd92c9..25e00a14ff 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 9ab23e1c4a..a100dbca3b 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_LOGICAL_PARALLEL_APPLY_MAIN,
-- 
2.25.1


--CE+1k2dSO48ffgeK
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v20-0002-Move-removal-of-old-serialized-snapshots-to-cust.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v5 1/8] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 214 ++++++++++++++++++++++++
 src/backend/postmaster/postmaster.c     |  44 ++++-
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  17 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 11 files changed, 301 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index dbbeac5a82..1b7aae60f5 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 0587e45920..7eae34884d 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..5f2b647544
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,214 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process is new as of Postgres 15.  It's main purpose is to
+ * offload tasks that could otherwise delay startup and checkpointing, but
+ * it needn't be restricted to just those things.  Offloaded tasks should
+ * not be synchronous (e.g., checkpointing shouldn't need to wait for the
+ * custodian to complete a task before proceeding).  Also, ensure that any
+ * offloaded tasks are either not required during single-user mode or are
+ * performed separately during single-user mode.
+ *
+ * The custodian is not an essential process and can shutdown quickly when
+ * requested.  The custodian will wake up approximately once every 5
+ * minutes to perform its tasks, but backends can (and should) set its
+ * latch to wake it up sooner.
+ *
+ * Normal termination is by SIGTERM, which instructs the bgwriter to
+ * exit(0).  Emergency termination is by SIGQUIT; like any backend, the
+ * custodian will simply abort and exit on SIGQUIT.
+ *
+ * If the custodian exits unexpectedly, the postmaster treats that the same
+ * as a backend crash: shared memory may be corrupted, so remaining
+ * backends should be killed by SIGQUIT and then a recovery cycle started.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include <time.h>
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "utils/memutils.h"
+
+#define CUSTODIAN_TIMEOUT_S (300)		/* 5 minutes */
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.
+	 *
+	 * You might wonder why this isn't coded as an infinite loop around a
+	 * PG_TRY construct.  The reason is that this is the bottom of the
+	 * exception stack, and so with PG_TRY there would be no exception handler
+	 * in force at all during the CATCH part.  By leaving the outermost setjmp
+	 * always active, we have at least some chance of recovering from an error
+	 * during error recovery.  (If we get into an infinite loop thereby, it
+	 * will soon be stopped by overflow of elog.c's internal state stack.)
+	 *
+	 * Note that we use sigsetjmp(..., 1), so that the prevailing signal mask
+	 * (to wit, BlockSig) will be restored when longjmp'ing to here.  Thus,
+	 * signals other than SIGQUIT will be blocked until we complete error
+	 * recovery.  It might seem that this policy makes the HOLD_INTERRUPS()
+	 * call redundant, but it is not since InterruptPending might be set
+	 * already.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		pgstat_report_wait_end();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		pg_time_t	start_time;
+		pg_time_t	end_time;
+		int			elapsed_secs;
+		int			cur_timeout;
+
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		start_time = (pg_time_t) time(NULL);
+
+		/* TODO: offloaded tasks go here */
+
+		/* Calculate how long to sleep */
+		end_time = (pg_time_t) time(NULL);
+		elapsed_secs = end_time - start_time;
+		if (elapsed_secs >= CUSTODIAN_TIMEOUT_S)
+			continue;			/* no sleep for us */
+		cur_timeout = CUSTODIAN_TIMEOUT_S - elapsed_secs;
+
+		(void) WaitLatch(MyLatch,
+						 WL_LATCH_SET | WL_TIMEOUT | WL_EXIT_ON_PM_DEATH,
+						 cur_timeout * 1000L /* convert to ms */ ,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 735fed490b..a867412268 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -251,6 +251,7 @@ bool		remove_temp_files_after_crash = true;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -557,6 +558,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1818,13 +1820,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2781,6 +2786,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3108,6 +3115,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3210,6 +3219,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3683,6 +3706,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(CheckpointerPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) CustodianPID)));
+		signal_child(CustodianPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3886,6 +3921,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3923,6 +3961,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -4016,6 +4055,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4221,6 +4261,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 90283f8a9f..1e693b69e5 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->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index 60972c3a75..e10cc2d82b 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index 0868e5a24f..8b52757ea6 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -274,6 +274,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_STARTUP:
 			backendDesc = "startup";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 0abc3ad540..71f522878e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -328,6 +328,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_STARTUP,
 	B_WAL_RECEIVER,
 	B_WAL_SENDER,
@@ -432,6 +433,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -444,6 +446,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..cf0a04ca6c
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,17 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a58888f9e9..ad61b4d802 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -374,11 +376,12 @@ 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 395d325c5f..1338d06823 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_PGSTAT_MAIN,
-- 
2.25.1


--2oS5YaxWCcQjTEyO
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v5-0002-Also-remove-pgsql_tmp-directories-during-startup.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v12 1/6] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 383 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  44 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 489 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..e90f5d0d1f
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,383 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(bool retry);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks(true);
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If retry is true, the custodian will re-enqueue the currently running task if
+ * an exception is encountered.
+ */
+static void
+DoCustodianTasks(bool retry)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (retry)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * If immediate is true, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  This is
+ * mostly useful for single-user mode.  If immediate is false, the task is added
+ * to the custodian's queue if it is not already enqueued, and this function
+ * returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, bool immediate, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (immediate)
+		DoCustodianTasks(false);
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 0b637ba6a2..3706eec25e 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -248,6 +248,7 @@ bool		remove_temp_files_after_crash = true;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -544,6 +545,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1821,13 +1823,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2746,6 +2751,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3066,6 +3073,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3159,6 +3168,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3616,6 +3639,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(CheckpointerPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) CustodianPID)));
+		signal_child(CustodianPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3793,6 +3828,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3830,6 +3868,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3919,6 +3958,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4113,6 +4153,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 13fa07b0ff..1bae34d1ee 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index 92f24a6c9b..d8e6ea45bc 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..170ca61a21
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, bool immediate, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 8d096fdeeb..448dde0161 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 6f2d5612e0..58455dc016 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--UlVJffcvxoiEqYs2
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v12-0002-Also-remove-pgsql_tmp-directories-during-startup.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v14 1/3] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 382 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 482 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..a94381bc21
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,382 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(void);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks();
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If we are not in a standalone backend, the custodian will re-enqueue the
+ * currently running task if an exception is encountered.
+ */
+static void
+DoCustodianTasks(void)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (IsPostmasterEnvironment)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * In standalone backends, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  Otherwise,
+ * the task is added to the custodian's queue if it is not already enqueued,
+ * and this function returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (!IsPostmasterEnvironment)
+		DoCustodianTasks();
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index a8a246921f..6a74423172 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -240,6 +240,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -537,6 +538,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1808,13 +1810,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2728,6 +2733,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3025,6 +3032,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3118,6 +3127,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3532,6 +3555,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3685,6 +3714,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3722,6 +3754,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3815,6 +3848,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4027,6 +4061,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index b1c35653fc..6a8485e865 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index b2abd75ddb..63fd242b1e 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..73d0bc5f02
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index aa13e1d66e..8f0e696663 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 0b2100be4a..48602c8a16 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--BXVAT5kNtrzKuDFl
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v14-0002-Move-removal-of-old-serialized-snapshots-to-cust.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v4 1/8] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 213 ++++++++++++++++++++++++
 src/backend/postmaster/postmaster.c     |  44 ++++-
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  17 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 11 files changed, 300 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index dbbeac5a82..1b7aae60f5 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 0587e45920..7eae34884d 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..dd86f0f5ce
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,213 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process is new as of Postgres 15.  It's main purpose is to
+ * offload tasks that could otherwise delay startup and checkpointing, but
+ * it needn't be restricted to just those things.  Offloaded tasks should
+ * not be synchronous (e.g., checkpointing shouldn't need to wait for the
+ * custodian to complete a task before proceeding).  Also, ensure that any
+ * offloaded tasks are either not required during single-user mode or are
+ * performed separately during single-user mode.
+ *
+ * The custodian is not an essential process and can shutdown quickly when
+ * requested.  The custodian will wake up approximately once every 5
+ * minutes to perform its tasks, but backends can (and should) set its
+ * latch to wake it up sooner.
+ *
+ * Normal termination is by SIGTERM, which instructs the bgwriter to
+ * exit(0).  Emergency termination is by SIGQUIT; like any backend, the
+ * custodian will simply abort and exit on SIGQUIT.
+ *
+ * If the custodian exits unexpectedly, the postmaster treats that the same
+ * as a backend crash: shared memory may be corrupted, so remaining
+ * backends should be killed by SIGQUIT and then a recovery cycle started.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include <time.h>
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "utils/memutils.h"
+
+#define CUSTODIAN_TIMEOUT_S (300)		/* 5 minutes */
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.
+	 *
+	 * You might wonder why this isn't coded as an infinite loop around a
+	 * PG_TRY construct.  The reason is that this is the bottom of the
+	 * exception stack, and so with PG_TRY there would be no exception handler
+	 * in force at all during the CATCH part.  By leaving the outermost setjmp
+	 * always active, we have at least some chance of recovering from an error
+	 * during error recovery.  (If we get into an infinite loop thereby, it
+	 * will soon be stopped by overflow of elog.c's internal state stack.)
+	 *
+	 * Note that we use sigsetjmp(..., 1), so that the prevailing signal mask
+	 * (to wit, BlockSig) will be restored when longjmp'ing to here.  Thus,
+	 * signals other than SIGQUIT will be blocked until we complete error
+	 * recovery.  It might seem that this policy makes the HOLD_INTERRUPS()
+	 * call redundant, but it is not since InterruptPending might be set
+	 * already.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		pgstat_report_wait_end();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		pg_time_t	start_time;
+		pg_time_t	end_time;
+		int			elapsed_secs;
+		int			cur_timeout;
+
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		start_time = (pg_time_t) time(NULL);
+
+		/* TODO: offloaded tasks go here */
+
+		/* Calculate how long to sleep */
+		end_time = (pg_time_t) time(NULL);
+		elapsed_secs = end_time - start_time;
+		if (elapsed_secs >= CUSTODIAN_TIMEOUT_S)
+			continue;			/* no sleep for us */
+		cur_timeout = CUSTODIAN_TIMEOUT_S - elapsed_secs;
+
+		(void) WaitLatch(MyLatch,
+						 WL_LATCH_SET | WL_TIMEOUT | WL_EXIT_ON_PM_DEATH,
+						 cur_timeout * 1000L /* convert to ms */ ,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index ce90877154..0911127471 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -250,6 +250,7 @@ bool		remove_temp_files_after_crash = true;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -556,6 +557,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1817,13 +1819,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2780,6 +2785,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3107,6 +3114,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3209,6 +3218,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3682,6 +3705,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(CheckpointerPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) CustodianPID)));
+		signal_child(CustodianPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3885,6 +3920,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3922,6 +3960,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -4015,6 +4054,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4220,6 +4260,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 37f032e7b9..f9df0259fd 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->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index 60972c3a75..e10cc2d82b 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index 0868e5a24f..8b52757ea6 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -274,6 +274,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_STARTUP:
 			backendDesc = "startup";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 0abc3ad540..71f522878e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -328,6 +328,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_STARTUP,
 	B_WAL_RECEIVER,
 	B_WAL_SENDER,
@@ -432,6 +433,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -444,6 +446,7 @@ extern AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..cf0a04ca6c
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,17 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index a58888f9e9..ad61b4d802 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -357,6 +357,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -374,11 +376,12 @@ 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 395d325c5f..1338d06823 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_PGSTAT_MAIN,
-- 
2.25.1


--BXVAT5kNtrzKuDFl
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v4-0002-Also-remove-pgsql_tmp-directories-during-startup.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v17 1/4] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 382 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 482 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..a94381bc21
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,382 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(void);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks();
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If we are not in a standalone backend, the custodian will re-enqueue the
+ * currently running task if an exception is encountered.
+ */
+static void
+DoCustodianTasks(void)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (IsPostmasterEnvironment)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * In standalone backends, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  Otherwise,
+ * the task is added to the custodian's queue if it is not already enqueued,
+ * and this function returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (!IsPostmasterEnvironment)
+		DoCustodianTasks();
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index a8a246921f..6a74423172 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -240,6 +240,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -537,6 +538,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1808,13 +1810,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2728,6 +2733,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3025,6 +3032,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3118,6 +3127,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3532,6 +3555,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3685,6 +3714,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3722,6 +3754,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3815,6 +3848,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4027,6 +4061,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index b1c35653fc..6a8485e865 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index b2abd75ddb..63fd242b1e 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..73d0bc5f02
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index aa13e1d66e..8f0e696663 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 0b2100be4a..48602c8a16 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--ew6BAiZeqk4r7MaW
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v17-0002-Move-removal-of-old-serialized-snapshots-to-cust.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v11 1/6] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 383 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  44 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 489 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..e90f5d0d1f
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,383 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(bool retry);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks(true);
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If retry is true, the custodian will re-enqueue the currently running task if
+ * an exception is encountered.
+ */
+static void
+DoCustodianTasks(bool retry)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (retry)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * If immediate is true, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  This is
+ * mostly useful for single-user mode.  If immediate is false, the task is added
+ * to the custodian's queue if it is not already enqueued, and this function
+ * returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, bool immediate, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (immediate)
+		DoCustodianTasks(false);
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 383bc4776e..b1b249cc90 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -248,6 +248,7 @@ bool		remove_temp_files_after_crash = true;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -544,6 +545,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1821,13 +1823,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2746,6 +2751,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3066,6 +3073,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3159,6 +3168,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3616,6 +3639,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(CheckpointerPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) CustodianPID)));
+		signal_child(CustodianPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3793,6 +3828,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3830,6 +3868,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3919,6 +3958,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4113,6 +4153,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 37aaab1338..f297f489c9 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index 92f24a6c9b..d8e6ea45bc 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index 683f616b1a..0131862973 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index ee48e392ed..c2e9bb3a75 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -426,6 +427,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -438,6 +440,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..170ca61a21
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, bool immediate, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 91824b4691..86acd3a5b9 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -396,6 +396,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -413,11 +415,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 6f2d5612e0..58455dc016 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--envbJBWh7q8WU6mo
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v11-0002-Also-remove-pgsql_tmp-directories-during-startup.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v9 1/6] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 383 ++++++++++++++++++++++++
 src/backend/postmaster/postmaster.c     |  44 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 12 files changed, 488 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..e90f5d0d1f
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,383 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(bool retry);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks(true);
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If retry is true, the custodian will re-enqueue the currently running task if
+ * an exception is encountered.
+ */
+static void
+DoCustodianTasks(bool retry)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (retry)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * If immediate is true, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  This is
+ * mostly useful for single-user mode.  If immediate is false, the task is added
+ * to the custodian's queue if it is not already enqueued, and this function
+ * returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, bool immediate, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (immediate)
+		DoCustodianTasks(false);
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 1664fcee2a..b25c180886 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -248,6 +248,7 @@ bool		remove_temp_files_after_crash = true;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -544,6 +545,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1821,13 +1823,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2750,6 +2755,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3070,6 +3077,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3163,6 +3172,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3620,6 +3643,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(CheckpointerPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) CustodianPID)));
+		signal_child(CustodianPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3797,6 +3832,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3834,6 +3872,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3923,6 +3962,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4117,6 +4157,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 1a6f527051..b19d743cab 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -129,6 +130,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -277,6 +279,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 37aaab1338..f297f489c9 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index 92f24a6c9b..d8e6ea45bc 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index 683f616b1a..0131862973 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 65cf4ba50f..36a83018e2 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -426,6 +427,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -438,6 +440,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..170ca61a21
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, bool immediate, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 2579e619eb..467421e371 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -394,6 +394,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -411,11 +413,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 6f2d5612e0..58455dc016 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--y0ulUmNC+osPPQO6
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v9-0002-Also-remove-pgsql_tmp-directories-during-startup.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v15 1/4] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 382 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 482 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..a94381bc21
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,382 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(void);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks();
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If we are not in a standalone backend, the custodian will re-enqueue the
+ * currently running task if an exception is encountered.
+ */
+static void
+DoCustodianTasks(void)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (IsPostmasterEnvironment)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * In standalone backends, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  Otherwise,
+ * the task is added to the custodian's queue if it is not already enqueued,
+ * and this function returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (!IsPostmasterEnvironment)
+		DoCustodianTasks();
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index a8a246921f..6a74423172 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -240,6 +240,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -537,6 +538,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1808,13 +1810,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2728,6 +2733,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3025,6 +3032,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3118,6 +3127,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3532,6 +3555,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3685,6 +3714,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3722,6 +3754,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3815,6 +3848,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4027,6 +4061,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index b1c35653fc..6a8485e865 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index b2abd75ddb..63fd242b1e 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..73d0bc5f02
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index aa13e1d66e..8f0e696663 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 0b2100be4a..48602c8a16 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--HcAYCG3uE/tztfnV
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v15-0002-Move-removal-of-old-serialized-snapshots-to-cust.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v10 1/6] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 383 ++++++++++++++++++++++++
 src/backend/postmaster/postmaster.c     |  44 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 12 files changed, 488 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..e90f5d0d1f
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,383 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(bool retry);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks(true);
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If retry is true, the custodian will re-enqueue the currently running task if
+ * an exception is encountered.
+ */
+static void
+DoCustodianTasks(bool retry)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (retry)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * If immediate is true, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  This is
+ * mostly useful for single-user mode.  If immediate is false, the task is added
+ * to the custodian's queue if it is not already enqueued, and this function
+ * returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, bool immediate, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (immediate)
+		DoCustodianTasks(false);
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 1664fcee2a..b25c180886 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -248,6 +248,7 @@ bool		remove_temp_files_after_crash = true;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -544,6 +545,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1821,13 +1823,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2750,6 +2755,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3070,6 +3077,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3163,6 +3172,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3620,6 +3643,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(CheckpointerPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) CustodianPID)));
+		signal_child(CustodianPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3797,6 +3832,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3834,6 +3872,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3923,6 +3962,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4117,6 +4157,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 1a6f527051..b19d743cab 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -129,6 +130,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -277,6 +279,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 37aaab1338..f297f489c9 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index 92f24a6c9b..d8e6ea45bc 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index 683f616b1a..0131862973 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 65cf4ba50f..36a83018e2 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -426,6 +427,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -438,6 +440,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..170ca61a21
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, bool immediate, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 2579e619eb..467421e371 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -394,6 +394,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -411,11 +413,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 6f2d5612e0..58455dc016 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--+QahgC5+KEYLbs62
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v10-0002-Also-remove-pgsql_tmp-directories-during-startup.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v13 1/6] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 383 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 483 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..e90f5d0d1f
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,383 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(bool retry);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks(true);
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If retry is true, the custodian will re-enqueue the currently running task if
+ * an exception is encountered.
+ */
+static void
+DoCustodianTasks(bool retry)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (retry)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * If immediate is true, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  This is
+ * mostly useful for single-user mode.  If immediate is false, the task is added
+ * to the custodian's queue if it is not already enqueued, and this function
+ * returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, bool immediate, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (immediate)
+		DoCustodianTasks(false);
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index c83cc8cc6c..00d18ee761 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -240,6 +240,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -537,6 +538,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1808,13 +1810,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2728,6 +2733,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3025,6 +3032,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3118,6 +3127,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3532,6 +3555,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3685,6 +3714,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3722,6 +3754,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3815,6 +3848,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4027,6 +4061,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index b1c35653fc..6a8485e865 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index b2abd75ddb..63fd242b1e 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..170ca61a21
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, bool immediate, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index aa13e1d66e..8f0e696663 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 0b2100be4a..48602c8a16 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--CE+1k2dSO48ffgeK
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v13-0002-Also-remove-pgsql_tmp-directories-during-startup.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v6 1/6] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 252 ++++++++++++++++++++++++
 src/backend/postmaster/postmaster.c     |  44 ++++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  20 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 12 files changed, 345 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 39ac4490db..620a0b1bae 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..db00282658
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,252 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  Also, ensure that any offloaded
+ * tasks are either not required during single-user mode or are performed
+ * separately during single-user mode.
+ *
+ * The custodian is not an essential process and can shutdown quickly when
+ * requested.  The custodian will wake up approximately once every 5
+ * minutes to perform its tasks, but backends can (and should) set its
+ * latch to wake it up sooner.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include <time.h>
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+#define CUSTODIAN_TIMEOUT_S (300)		/* 5 minutes */
+
+typedef struct
+{
+	slock_t		cust_lck;
+	int			cust_flags;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * On startup and after an exception, we won't know exactly what tasks need
+	 * to be performed, so request all of them.
+	 */
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+	CustodianShmem->cust_flags = 0xFFFFFFFF;
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		pg_time_t	start_time;
+		pg_time_t	end_time;
+		int			elapsed_secs;
+		int			cur_timeout;
+		int			flags;
+
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		start_time = (pg_time_t) time(NULL);
+
+		/* Obtain requested tasks */
+		SpinLockAcquire(&CustodianShmem->cust_lck);
+		flags = CustodianShmem->cust_flags;
+		CustodianShmem->cust_flags = 0;
+		SpinLockRelease(&CustodianShmem->cust_lck);
+
+		/* TODO: offloaded tasks go here */
+
+		/* Calculate how long to sleep */
+		end_time = (pg_time_t) time(NULL);
+		elapsed_secs = end_time - start_time;
+		if (elapsed_secs >= CUSTODIAN_TIMEOUT_S)
+			continue;			/* no sleep for us */
+		cur_timeout = CUSTODIAN_TIMEOUT_S - elapsed_secs;
+
+		(void) WaitLatch(MyLatch,
+						 WL_LATCH_SET | WL_TIMEOUT | WL_EXIT_ON_PM_DEATH,
+						 cur_timeout * 1000L /* convert to ms */ ,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ */
+void
+RequestCustodian(int flags)
+{
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+	CustodianShmem->cust_flags |= flags;
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index dde4bc25b1..5162ee9dec 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -251,6 +251,7 @@ bool		remove_temp_files_after_crash = true;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -548,6 +549,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1823,13 +1825,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2769,6 +2774,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3089,6 +3096,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3182,6 +3191,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3639,6 +3662,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(CheckpointerPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) CustodianPID)));
+		signal_child(CustodianPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3816,6 +3851,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3853,6 +3891,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3942,6 +3981,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4135,6 +4175,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 1a6f527051..b19d743cab 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -129,6 +130,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -277,6 +279,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 37aaab1338..f297f489c9 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index 87c15b9c6f..469768c4e4 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index b25bd0e583..66bf42e5b1 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -273,6 +273,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_STARTUP:
 			backendDesc = "startup";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 0af130fbc5..ffe9404c68 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -330,6 +330,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_STARTUP,
 	B_WAL_RECEIVER,
 	B_WAL_SENDER,
@@ -433,6 +434,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -445,6 +447,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..c95a7c7de6
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,20 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(int flags);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 2579e619eb..467421e371 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -394,6 +394,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -411,11 +413,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index b578e2ec75..7524e197e5 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--liOOAslEiF7prFVr
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v6-0002-Also-remove-pgsql_tmp-directories-during-startup.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v16 1/4] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 382 ++++++++++++++++++++++++
 src/backend/postmaster/meson.build      |   1 +
 src/backend/postmaster/postmaster.c     |  38 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 13 files changed, 482 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..a94381bc21
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,382 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(void);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks();
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If we are not in a standalone backend, the custodian will re-enqueue the
+ * currently running task if an exception is encountered.
+ */
+static void
+DoCustodianTasks(void)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (IsPostmasterEnvironment)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * In standalone backends, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  Otherwise,
+ * the task is added to the custodian's queue if it is not already enqueued,
+ * and this function returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (!IsPostmasterEnvironment)
+		DoCustodianTasks();
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/meson.build b/src/backend/postmaster/meson.build
index 293a44ca29..ac72a8a07f 100644
--- a/src/backend/postmaster/meson.build
+++ b/src/backend/postmaster/meson.build
@@ -4,6 +4,7 @@ backend_sources += files(
   'bgworker.c',
   'bgwriter.c',
   'checkpointer.c',
+  'custodian.c',
   'fork_process.c',
   'interrupt.c',
   'pgarch.c',
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index a8a246921f..6a74423172 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -240,6 +240,7 @@ bool		send_abort_for_kill = false;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -537,6 +538,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1808,13 +1810,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2728,6 +2733,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3025,6 +3032,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3118,6 +3127,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3532,6 +3555,12 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 	else if (CheckpointerPID != 0 && take_action)
 		sigquit_child(CheckpointerPID);
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+		sigquit_child(CustodianPID);
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3685,6 +3714,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3722,6 +3754,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3815,6 +3848,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4027,6 +4061,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index b204ecdbc3..cf80e65779 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -130,6 +131,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -278,6 +280,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index b1c35653fc..6a8485e865 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index b2abd75ddb..63fd242b1e 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index eb1046450b..f19f4c3075 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -278,6 +278,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_LOGGER:
 			backendDesc = "logger";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 795182fa51..59a95dd7c0 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -323,6 +323,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_LOGGER,
 	B_STANDALONE_BACKEND,
 	B_STARTUP,
@@ -429,6 +430,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -441,6 +443,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..73d0bc5f02
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index aa13e1d66e..8f0e696663 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -400,6 +400,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -417,11 +419,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 0b2100be4a..48602c8a16 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--lrZ03NoBR/3+SXJZ
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v16-0002-Move-removal-of-old-serialized-snapshots-to-cust.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v8 1/6] Introduce custodian.
@ 2022-01-05 19:24 Nathan Bossart <bossartn@amazon.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Nathan Bossart @ 2022-01-05 19:24 UTC (permalink / raw)

The custodian process is a new auxiliary process that is intended
to help offload tasks could otherwise delay startup and
checkpointing.  This commit simply adds the new process; it does
not yet do anything useful.
---
 src/backend/postmaster/Makefile         |   1 +
 src/backend/postmaster/auxprocess.c     |   8 +
 src/backend/postmaster/custodian.c      | 383 ++++++++++++++++++++++++
 src/backend/postmaster/postmaster.c     |  44 ++-
 src/backend/storage/ipc/ipci.c          |   3 +
 src/backend/storage/lmgr/proc.c         |   1 +
 src/backend/utils/activity/wait_event.c |   3 +
 src/backend/utils/init/miscinit.c       |   3 +
 src/include/miscadmin.h                 |   3 +
 src/include/postmaster/custodian.h      |  32 ++
 src/include/storage/proc.h              |  11 +-
 src/include/utils/wait_event.h          |   1 +
 12 files changed, 488 insertions(+), 5 deletions(-)
 create mode 100644 src/backend/postmaster/custodian.c
 create mode 100644 src/include/postmaster/custodian.h

diff --git a/src/backend/postmaster/Makefile b/src/backend/postmaster/Makefile
index 3a794e54d6..e1e1d1123f 100644
--- a/src/backend/postmaster/Makefile
+++ b/src/backend/postmaster/Makefile
@@ -18,6 +18,7 @@ OBJS = \
 	bgworker.o \
 	bgwriter.o \
 	checkpointer.o \
+	custodian.o \
 	fork_process.o \
 	interrupt.o \
 	pgarch.o \
diff --git a/src/backend/postmaster/auxprocess.c b/src/backend/postmaster/auxprocess.c
index 7765d1c83d..c275271c95 100644
--- a/src/backend/postmaster/auxprocess.c
+++ b/src/backend/postmaster/auxprocess.c
@@ -20,6 +20,7 @@
 #include "pgstat.h"
 #include "postmaster/auxprocess.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/startup.h"
 #include "postmaster/walwriter.h"
 #include "replication/walreceiver.h"
@@ -74,6 +75,9 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 		case CheckpointerProcess:
 			MyBackendType = B_CHECKPOINTER;
 			break;
+		case CustodianProcess:
+			MyBackendType = B_CUSTODIAN;
+			break;
 		case WalWriterProcess:
 			MyBackendType = B_WAL_WRITER;
 			break;
@@ -153,6 +157,10 @@ AuxiliaryProcessMain(AuxProcType auxtype)
 			CheckpointerMain();
 			proc_exit(1);
 
+		case CustodianProcess:
+			CustodianMain();
+			proc_exit(1);
+
 		case WalWriterProcess:
 			WalWriterMain();
 			proc_exit(1);
diff --git a/src/backend/postmaster/custodian.c b/src/backend/postmaster/custodian.c
new file mode 100644
index 0000000000..e90f5d0d1f
--- /dev/null
+++ b/src/backend/postmaster/custodian.c
@@ -0,0 +1,383 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.c
+ *
+ * The custodian process handles a variety of non-critical tasks that might
+ * otherwise delay startup, checkpointing, etc.  Offloaded tasks should not
+ * be synchronous (e.g., checkpointing shouldn't wait for the custodian to
+ * complete a task before proceeding).  However, tasks can be synchronously
+ * executed when necessary (e.g., single-user mode).  The custodian is not
+ * an essential process and can shutdown quickly when requested.  The
+ * custodian only wakes up to perform its tasks when its latch is set.
+ *
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *   src/backend/postmaster/custodian.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "libpq/pqsignal.h"
+#include "pgstat.h"
+#include "postmaster/custodian.h"
+#include "postmaster/interrupt.h"
+#include "storage/bufmgr.h"
+#include "storage/condition_variable.h"
+#include "storage/fd.h"
+#include "storage/proc.h"
+#include "storage/procsignal.h"
+#include "storage/smgr.h"
+#include "utils/memutils.h"
+
+static void DoCustodianTasks(bool retry);
+static CustodianTask CustodianGetNextTask(void);
+static void CustodianEnqueueTask(CustodianTask task);
+static const struct cust_task_funcs_entry *LookupCustodianFunctions(CustodianTask task);
+
+typedef struct
+{
+	slock_t		cust_lck;
+
+	CustodianTask task_queue_elems[NUM_CUSTODIAN_TASKS];
+	int			task_queue_head;
+} CustodianShmemStruct;
+
+static CustodianShmemStruct *CustodianShmem;
+
+typedef void (*CustodianTaskFunction) (void);
+typedef void (*CustodianTaskHandleArg) (Datum arg);
+
+struct cust_task_funcs_entry
+{
+	CustodianTask task;
+	CustodianTaskFunction task_func;		/* performs task */
+	CustodianTaskHandleArg handle_arg_func;	/* handles additional info in request */
+};
+
+/*
+ * Add new tasks here.
+ *
+ * task_func is the logic that will be executed via DoCustodianTasks() when the
+ * matching task is requested via RequestCustodian().  handle_arg_func is an
+ * optional function for providing extra information for the next invocation of
+ * the task.  Typically, the extra information should be stored in shared
+ * memory for access from the custodian process.  handle_arg_func is invoked
+ * before enqueueing the task, and it will still be invoked regardless of
+ * whether the task is already enqueued.
+ */
+static const struct cust_task_funcs_entry cust_task_functions[] = {
+	{INVALID_CUSTODIAN_TASK, NULL, NULL}	/* must be last */
+};
+
+/*
+ * Main entry point for custodian process
+ *
+ * This is invoked from AuxiliaryProcessMain, which has already created the
+ * basic execution environment, but not enabled signals yet.
+ */
+void
+CustodianMain(void)
+{
+	sigjmp_buf	local_sigjmp_buf;
+	MemoryContext custodian_context;
+
+	/*
+	 * Properly accept or ignore signals that might be sent to us.
+	 */
+	pqsignal(SIGHUP, SignalHandlerForConfigReload);
+	pqsignal(SIGINT, SignalHandlerForShutdownRequest);
+	pqsignal(SIGTERM, SignalHandlerForShutdownRequest);
+	/* SIGQUIT handler was already set up by InitPostmasterChild */
+	pqsignal(SIGALRM, SIG_IGN);
+	pqsignal(SIGPIPE, SIG_IGN);
+	pqsignal(SIGUSR1, procsignal_sigusr1_handler);
+	pqsignal(SIGUSR2, SIG_IGN);
+
+	/*
+	 * Reset some signals that are accepted by postmaster but not here
+	 */
+	pqsignal(SIGCHLD, SIG_DFL);
+
+	/*
+	 * Create a memory context that we will do all our work in.  We do this so
+	 * that we can reset the context during error recovery and thereby avoid
+	 * possible memory leaks.
+	 */
+	custodian_context = AllocSetContextCreate(TopMemoryContext,
+											  "Custodian",
+											  ALLOCSET_DEFAULT_SIZES);
+	MemoryContextSwitchTo(custodian_context);
+
+	/*
+	 * If an exception is encountered, processing resumes here.  As with other
+	 * auxiliary processes, we cannot use PG_TRY because this is the bottom of
+	 * the exception stack.
+	 */
+	if (sigsetjmp(local_sigjmp_buf, 1) != 0)
+	{
+		/* Since not using PG_TRY, must reset error stack by hand */
+		error_context_stack = NULL;
+
+		/* Prevent interrupts while cleaning up */
+		HOLD_INTERRUPTS();
+
+		/* Report the error to the server log */
+		EmitErrorReport();
+
+		/*
+		 * These operations are really just a minimal subset of
+		 * AbortTransaction().  We don't have very many resources to worry
+		 * about.
+		 */
+		LWLockReleaseAll();
+		ConditionVariableCancelSleep();
+		AbortBufferIO();
+		UnlockBuffers();
+		ReleaseAuxProcessResources(false);
+		AtEOXact_Buffers(false);
+		AtEOXact_SMgr();
+		AtEOXact_Files(false);
+		AtEOXact_HashTables(false);
+
+		/*
+		 * Now return to normal top-level context and clear ErrorContext for
+		 * next time.
+		 */
+		MemoryContextSwitchTo(custodian_context);
+		FlushErrorState();
+
+		/* Flush any leaked data in the top-level context */
+		MemoryContextResetAndDeleteChildren(custodian_context);
+
+		/* Now we can allow interrupts again */
+		RESUME_INTERRUPTS();
+
+		/*
+		 * Sleep at least 1 second after any error.  A write error is likely
+		 * to be repeated, and we don't want to be filling the error logs as
+		 * fast as we can.
+		 */
+		pg_usleep(1000000L);
+
+		/*
+		 * Close all open files after any error.  This is helpful on Windows,
+		 * where holding deleted files open causes various strange errors.
+		 * It's not clear we need it elsewhere, but shouldn't hurt.
+		 */
+		smgrcloseall();
+
+		/* Report wait end here, when there is no further possibility of wait */
+		pgstat_report_wait_end();
+	}
+
+	/* We can now handle ereport(ERROR) */
+	PG_exception_stack = &local_sigjmp_buf;
+
+	/*
+	 * Unblock signals (they were blocked when the postmaster forked us)
+	 */
+	PG_SETMASK(&UnBlockSig);
+
+	/*
+	 * Advertise out latch that backends can use to wake us up while we're
+	 * sleeping.
+	 */
+	ProcGlobal->custodianLatch = &MyProc->procLatch;
+
+	/*
+	 * Loop forever
+	 */
+	for (;;)
+	{
+		/* Clear any already-pending wakeups */
+		ResetLatch(MyLatch);
+
+		HandleMainLoopInterrupts();
+
+		DoCustodianTasks(true);
+
+		(void) WaitLatch(MyLatch, WL_LATCH_SET | WL_EXIT_ON_PM_DEATH, 0,
+						 WAIT_EVENT_CUSTODIAN_MAIN);
+	}
+
+	pg_unreachable();
+}
+
+/*
+ * DoCustodianTasks
+ *		Perform requested custodian tasks
+ *
+ * If retry is true, the custodian will re-enqueue the currently running task if
+ * an exception is encountered.
+ */
+static void
+DoCustodianTasks(bool retry)
+{
+	CustodianTask	task;
+
+	while ((task = CustodianGetNextTask()) != INVALID_CUSTODIAN_TASK)
+	{
+		CustodianTaskFunction func = (LookupCustodianFunctions(task))->task_func;
+
+		PG_TRY();
+		{
+			(*func) ();
+		}
+		PG_CATCH();
+		{
+			if (retry)
+				CustodianEnqueueTask(task);
+
+			PG_RE_THROW();
+		}
+		PG_END_TRY();
+	}
+}
+
+Size
+CustodianShmemSize(void)
+{
+	return sizeof(CustodianShmemStruct);
+}
+
+void
+CustodianShmemInit(void)
+{
+	Size		size = CustodianShmemSize();
+	bool		found;
+
+	CustodianShmem = (CustodianShmemStruct *)
+		ShmemInitStruct("Custodian Data", size, &found);
+
+	if (!found)
+	{
+		memset(CustodianShmem, 0, size);
+		SpinLockInit(&CustodianShmem->cust_lck);
+		for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+			CustodianShmem->task_queue_elems[i] = INVALID_CUSTODIAN_TASK;
+	}
+}
+
+/*
+ * RequestCustodian
+ *		Called to request a custodian task.
+ *
+ * If immediate is true, the task is performed immediately in the current
+ * process, and this function will not return until it completes.  This is
+ * mostly useful for single-user mode.  If immediate is false, the task is added
+ * to the custodian's queue if it is not already enqueued, and this function
+ * returns without waiting for the task to complete.
+ *
+ * arg can be used to provide additional information to the custodian that is
+ * necessary for the task.  Typically, the handling function should store this
+ * information in shared memory for later use by the custodian.  Note that the
+ * task's handling function for arg is invoked before enqueueing the task, and
+ * it will still be invoked regardless of whether the task is already enqueued.
+ */
+void
+RequestCustodian(CustodianTask requested, bool immediate, Datum arg)
+{
+	CustodianTaskHandleArg arg_func = (LookupCustodianFunctions(requested))->handle_arg_func;
+
+	/* First process any extra information provided in the request. */
+	if (arg_func)
+		(*arg_func) (arg);
+
+	CustodianEnqueueTask(requested);
+
+	if (immediate)
+		DoCustodianTasks(false);
+	else if (ProcGlobal->custodianLatch)
+		SetLatch(ProcGlobal->custodianLatch);
+}
+
+/*
+ * CustodianEnqueueTask
+ *		Add a task to the custodian's queue
+ *
+ * If the task is already in the queue, this function has no effect.
+ */
+static void
+CustodianEnqueueTask(CustodianTask task)
+{
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	for (int i = 0; i < NUM_CUSTODIAN_TASKS; i++)
+	{
+		int idx = (CustodianShmem->task_queue_head + i) % NUM_CUSTODIAN_TASKS;
+		CustodianTask *elem = &CustodianShmem->task_queue_elems[idx];
+
+		/*
+		 * If the task is already queued in this slot or the slot is empty,
+		 * enqueue the task here and return.
+		 */
+		if (*elem == INVALID_CUSTODIAN_TASK || *elem == task)
+		{
+			*elem = task;
+			SpinLockRelease(&CustodianShmem->cust_lck);
+			return;
+		}
+	}
+
+	/* We should never run out of space in the queue. */
+	elog(ERROR, "could not enqueue custodian task %d", task);
+	pg_unreachable();
+}
+
+/*
+ * CustodianGetNextTask
+ *		Retrieve the next task that the custodian should execute
+ *
+ * The returned task is dequeued from the custodian's queue.  If no tasks are
+ * queued, INVALID_CUSTODIAN_TASK is returned.
+ */
+static CustodianTask
+CustodianGetNextTask(void)
+{
+	CustodianTask next_task;
+	CustodianTask *elem;
+
+	SpinLockAcquire(&CustodianShmem->cust_lck);
+
+	elem = &CustodianShmem->task_queue_elems[CustodianShmem->task_queue_head];
+
+	next_task = *elem;
+	*elem = INVALID_CUSTODIAN_TASK;
+
+	CustodianShmem->task_queue_head++;
+	CustodianShmem->task_queue_head %= NUM_CUSTODIAN_TASKS;
+
+	SpinLockRelease(&CustodianShmem->cust_lck);
+
+	return next_task;
+}
+
+/*
+ * LookupCustodianFunctions
+ *		Given a custodian task, look up its function pointers.
+ */
+static const struct cust_task_funcs_entry *
+LookupCustodianFunctions(CustodianTask task)
+{
+	const struct cust_task_funcs_entry *entry;
+
+	Assert(task >= 0 && task < NUM_CUSTODIAN_TASKS);
+
+	for (entry = cust_task_functions;
+		 entry && entry->task != INVALID_CUSTODIAN_TASK;
+		 entry++)
+	{
+		if (entry->task == task)
+			return entry;
+	}
+
+	/* All tasks must have an entry. */
+	elog(ERROR, "could not lookup functions for custodian task %d", task);
+	pg_unreachable();
+}
diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 81cb585891..d705ff6bf0 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -251,6 +251,7 @@ bool		remove_temp_files_after_crash = true;
 static pid_t StartupPID = 0,
 			BgWriterPID = 0,
 			CheckpointerPID = 0,
+			CustodianPID = 0,
 			WalWriterPID = 0,
 			WalReceiverPID = 0,
 			AutoVacPID = 0,
@@ -547,6 +548,7 @@ static void ShmemBackendArrayRemove(Backend *bn);
 #define StartArchiver()			StartChildProcess(ArchiverProcess)
 #define StartBackgroundWriter() StartChildProcess(BgWriterProcess)
 #define StartCheckpointer()		StartChildProcess(CheckpointerProcess)
+#define StartCustodian()		StartChildProcess(CustodianProcess)
 #define StartWalWriter()		StartChildProcess(WalWriterProcess)
 #define StartWalReceiver()		StartChildProcess(WalReceiverProcess)
 
@@ -1826,13 +1828,16 @@ ServerLoop(void)
 		/*
 		 * If no background writer process is running, and we are not in a
 		 * state that prevents it, start one.  It doesn't matter if this
-		 * fails, we'll just try again later.  Likewise for the checkpointer.
+		 * fails, we'll just try again later.  Likewise for the checkpointer
+		 * and custodian.
 		 */
 		if (pmState == PM_RUN || pmState == PM_RECOVERY ||
 			pmState == PM_HOT_STANDBY || pmState == PM_STARTUP)
 		{
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 		}
@@ -2755,6 +2760,8 @@ SIGHUP_handler(SIGNAL_ARGS)
 			signal_child(BgWriterPID, SIGHUP);
 		if (CheckpointerPID != 0)
 			signal_child(CheckpointerPID, SIGHUP);
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGHUP);
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGHUP);
 		if (WalReceiverPID != 0)
@@ -3075,6 +3082,8 @@ reaper(SIGNAL_ARGS)
 			 */
 			if (CheckpointerPID == 0)
 				CheckpointerPID = StartCheckpointer();
+			if (CustodianPID == 0)
+				CustodianPID = StartCustodian();
 			if (BgWriterPID == 0)
 				BgWriterPID = StartBackgroundWriter();
 			if (WalWriterPID == 0)
@@ -3168,6 +3177,20 @@ reaper(SIGNAL_ARGS)
 			continue;
 		}
 
+		/*
+		 * Was it the custodian?  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 == CustodianPID)
+		{
+			CustodianPID = 0;
+			if (!EXIT_STATUS_0(exitstatus))
+				HandleChildCrash(pid, exitstatus,
+								 _("custodian process"));
+			continue;
+		}
+
 		/*
 		 * Was it the wal writer?  Normal exit can be ignored; we'll start a
 		 * new one at the next iteration of the postmaster's main loop, if
@@ -3625,6 +3648,18 @@ HandleChildCrash(int pid, int exitstatus, const char *procname)
 		signal_child(CheckpointerPID, (SendStop ? SIGSTOP : SIGQUIT));
 	}
 
+	/* Take care of the custodian too */
+	if (pid == CustodianPID)
+		CustodianPID = 0;
+	else if (CustodianPID != 0 && take_action)
+	{
+		ereport(DEBUG2,
+				(errmsg_internal("sending %s to process %d",
+								 (SendStop ? "SIGSTOP" : "SIGQUIT"),
+								 (int) CustodianPID)));
+		signal_child(CustodianPID, (SendStop ? SIGSTOP : SIGQUIT));
+	}
+
 	/* Take care of the walwriter too */
 	if (pid == WalWriterPID)
 		WalWriterPID = 0;
@@ -3802,6 +3837,9 @@ PostmasterStateMachine(void)
 		/* and the bgwriter too */
 		if (BgWriterPID != 0)
 			signal_child(BgWriterPID, SIGTERM);
+		/* and the custodian too */
+		if (CustodianPID != 0)
+			signal_child(CustodianPID, SIGTERM);
 		/* and the walwriter too */
 		if (WalWriterPID != 0)
 			signal_child(WalWriterPID, SIGTERM);
@@ -3839,6 +3877,7 @@ PostmasterStateMachine(void)
 			BgWriterPID == 0 &&
 			(CheckpointerPID == 0 ||
 			 (!FatalError && Shutdown < ImmediateShutdown)) &&
+			CustodianPID == 0 &&
 			WalWriterPID == 0 &&
 			AutoVacPID == 0)
 		{
@@ -3928,6 +3967,7 @@ PostmasterStateMachine(void)
 			Assert(WalReceiverPID == 0);
 			Assert(BgWriterPID == 0);
 			Assert(CheckpointerPID == 0);
+			Assert(CustodianPID == 0);
 			Assert(WalWriterPID == 0);
 			Assert(AutoVacPID == 0);
 			/* syslogger is not considered here */
@@ -4122,6 +4162,8 @@ TerminateChildren(int signal)
 		signal_child(BgWriterPID, signal);
 	if (CheckpointerPID != 0)
 		signal_child(CheckpointerPID, signal);
+	if (CustodianPID != 0)
+		signal_child(CustodianPID, signal);
 	if (WalWriterPID != 0)
 		signal_child(WalWriterPID, signal);
 	if (WalReceiverPID != 0)
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 1a6f527051..b19d743cab 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -30,6 +30,7 @@
 #include "postmaster/autovacuum.h"
 #include "postmaster/bgworker_internals.h"
 #include "postmaster/bgwriter.h"
+#include "postmaster/custodian.h"
 #include "postmaster/postmaster.h"
 #include "replication/logicallauncher.h"
 #include "replication/origin.h"
@@ -129,6 +130,7 @@ CalculateShmemSize(int *num_semaphores)
 	size = add_size(size, PMSignalShmemSize());
 	size = add_size(size, ProcSignalShmemSize());
 	size = add_size(size, CheckpointerShmemSize());
+	size = add_size(size, CustodianShmemSize());
 	size = add_size(size, AutoVacuumShmemSize());
 	size = add_size(size, ReplicationSlotsShmemSize());
 	size = add_size(size, ReplicationOriginShmemSize());
@@ -277,6 +279,7 @@ CreateSharedMemoryAndSemaphores(void)
 	PMSignalShmemInit();
 	ProcSignalShmemInit();
 	CheckpointerShmemInit();
+	CustodianShmemInit();
 	AutoVacuumShmemInit();
 	ReplicationSlotsShmemInit();
 	ReplicationOriginShmemInit();
diff --git a/src/backend/storage/lmgr/proc.c b/src/backend/storage/lmgr/proc.c
index 37aaab1338..f297f489c9 100644
--- a/src/backend/storage/lmgr/proc.c
+++ b/src/backend/storage/lmgr/proc.c
@@ -180,6 +180,7 @@ InitProcGlobal(void)
 	ProcGlobal->startupBufferPinWaitBufId = -1;
 	ProcGlobal->walwriterLatch = NULL;
 	ProcGlobal->checkpointerLatch = NULL;
+	ProcGlobal->custodianLatch = NULL;
 	pg_atomic_init_u32(&ProcGlobal->procArrayGroupFirst, INVALID_PGPROCNO);
 	pg_atomic_init_u32(&ProcGlobal->clogGroupFirst, INVALID_PGPROCNO);
 
diff --git a/src/backend/utils/activity/wait_event.c b/src/backend/utils/activity/wait_event.c
index 92f24a6c9b..d8e6ea45bc 100644
--- a/src/backend/utils/activity/wait_event.c
+++ b/src/backend/utils/activity/wait_event.c
@@ -224,6 +224,9 @@ pgstat_get_wait_activity(WaitEventActivity w)
 		case WAIT_EVENT_CHECKPOINTER_MAIN:
 			event_name = "CheckpointerMain";
 			break;
+		case WAIT_EVENT_CUSTODIAN_MAIN:
+			event_name = "CustodianMain";
+			break;
 		case WAIT_EVENT_LOGICAL_APPLY_MAIN:
 			event_name = "LogicalApplyMain";
 			break;
diff --git a/src/backend/utils/init/miscinit.c b/src/backend/utils/init/miscinit.c
index bd973ba613..22037f0d99 100644
--- a/src/backend/utils/init/miscinit.c
+++ b/src/backend/utils/init/miscinit.c
@@ -273,6 +273,9 @@ GetBackendTypeDesc(BackendType backendType)
 		case B_CHECKPOINTER:
 			backendDesc = "checkpointer";
 			break;
+		case B_CUSTODIAN:
+			backendDesc = "custodian";
+			break;
 		case B_STARTUP:
 			backendDesc = "startup";
 			break;
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 067b729d5a..ffd59616b7 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -322,6 +322,7 @@ typedef enum BackendType
 	B_BG_WORKER,
 	B_BG_WRITER,
 	B_CHECKPOINTER,
+	B_CUSTODIAN,
 	B_STARTUP,
 	B_WAL_RECEIVER,
 	B_WAL_SENDER,
@@ -425,6 +426,7 @@ typedef enum
 	BgWriterProcess,
 	ArchiverProcess,
 	CheckpointerProcess,
+	CustodianProcess,
 	WalWriterProcess,
 	WalReceiverProcess,
 
@@ -437,6 +439,7 @@ extern PGDLLIMPORT AuxProcType MyAuxProcType;
 #define AmBackgroundWriterProcess() (MyAuxProcType == BgWriterProcess)
 #define AmArchiverProcess()			(MyAuxProcType == ArchiverProcess)
 #define AmCheckpointerProcess()		(MyAuxProcType == CheckpointerProcess)
+#define AmCustodianProcess()		(MyAuxProcType == CustodianProcess)
 #define AmWalWriterProcess()		(MyAuxProcType == WalWriterProcess)
 #define AmWalReceiverProcess()		(MyAuxProcType == WalReceiverProcess)
 
diff --git a/src/include/postmaster/custodian.h b/src/include/postmaster/custodian.h
new file mode 100644
index 0000000000..170ca61a21
--- /dev/null
+++ b/src/include/postmaster/custodian.h
@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * custodian.h
+ *   Exports from postmaster/custodian.c.
+ *
+ * Copyright (c) 2022, PostgreSQL Global Development Group
+ *
+ * src/include/postmaster/custodian.h
+ *
+ *-------------------------------------------------------------------------
+ */
+#ifndef _CUSTODIAN_H
+#define _CUSTODIAN_H
+
+/*
+ * If you add a new task here, be sure to add its corresponding function
+ * pointers to cust_task_functions in custodian.c.
+ */
+typedef enum CustodianTask
+{
+	FAKE_TASK,						/* placeholder until we have a real task */
+
+	NUM_CUSTODIAN_TASKS,			/* new tasks go above */
+	INVALID_CUSTODIAN_TASK
+} CustodianTask;
+
+extern void CustodianMain(void) pg_attribute_noreturn();
+extern Size CustodianShmemSize(void);
+extern void CustodianShmemInit(void);
+extern void RequestCustodian(CustodianTask task, bool immediate, Datum arg);
+
+#endif						/* _CUSTODIAN_H */
diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h
index 2579e619eb..467421e371 100644
--- a/src/include/storage/proc.h
+++ b/src/include/storage/proc.h
@@ -394,6 +394,8 @@ typedef struct PROC_HDR
 	Latch	   *walwriterLatch;
 	/* Checkpointer process's latch */
 	Latch	   *checkpointerLatch;
+	/* Custodian process's latch */
+	Latch	   *custodianLatch;
 	/* Current shared estimate of appropriate spins_per_delay value */
 	int			spins_per_delay;
 	/* Buffer id of the buffer that Startup process waits for pin on, or -1 */
@@ -411,11 +413,12 @@ extern PGDLLIMPORT 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, 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.
+ * Background writer, checkpointer, custodian, 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 6 slots.
  */
-#define NUM_AUXILIARY_PROCS		5
+#define NUM_AUXILIARY_PROCS		6
 
 /* configurable options */
 extern PGDLLIMPORT int DeadlockTimeout;
diff --git a/src/include/utils/wait_event.h b/src/include/utils/wait_event.h
index 6f2d5612e0..58455dc016 100644
--- a/src/include/utils/wait_event.h
+++ b/src/include/utils/wait_event.h
@@ -40,6 +40,7 @@ typedef enum
 	WAIT_EVENT_BGWRITER_HIBERNATE,
 	WAIT_EVENT_BGWRITER_MAIN,
 	WAIT_EVENT_CHECKPOINTER_MAIN,
+	WAIT_EVENT_CUSTODIAN_MAIN,
 	WAIT_EVENT_LOGICAL_APPLY_MAIN,
 	WAIT_EVENT_LOGICAL_LAUNCHER_MAIN,
 	WAIT_EVENT_RECOVERY_WAL_STREAM,
-- 
2.25.1


--X1bOJ3K7DJ5YkBrT
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment;
	filename="v8-0002-Also-remove-pgsql_tmp-directories-during-startup.patch"



^ permalink  raw  reply  [nested|flat] 24+ messages in thread

* [PATCH v7 3/3] handle relation statistics correctly during rewrites
@ 2025-11-04 13:52 Bertrand Drouvot <bertranddrouvot.pg@gmail.com>
  0 siblings, 0 replies; 24+ messages in thread

From: Bertrand Drouvot @ 2025-11-04 13:52 UTC (permalink / raw)

Now that PGSTAT_KIND_RELATION is keyed by refilenode, we need to handle rewrites.

To do so, this patch:

- Adds PgStat_PendingRewrite, a new struct to track rewrite operations within
a transaction, storing the old locator, new locator, and original locator (for
rewrite chains). This allows stats to be copied from the original location to
the final location at commit time.

- Adds a new function, pgstat_mark_rewrite(), called when a table rewrite begins.
It records the rewrite operation in a local list and detects rewrite chains by
checking if the old_locator matches any existing new_locator, preserving the
chain's original_locator.

- Modifies pgstat_copy_relation_stats(), to accept RelFileLocators instead of
Relations, with a new increment parameter to accumulate stats (needed for rewrite
chains with DML between rewrites).

- Ensures that AtEOXact_PgStat_Relations(), AtPrepare_PgStat_Relations(),
pgstat_twophase_postcommit()/postabort() pgstat_drop_relation() handle the
PgStat_PendingRewrite list correctly.

Note that due to the new flush call in pgstat_twophase_postcommit() we can not
call GetCurrentTransactionStopTimestamp() in pgstat_relation_flush_cb(). So,
adding a check to handle this special case and call GetCurrentTimestamp() instead.
Note that we'd call GetCurrentTimestamp() only if there is a rewrite, so that
the GetCurrentTimestamp() extra cost should be negligible. Another solution
could be to trigger the flush from FinishPreparedTransaction() but that's not
worth the extra complexity.

The new pending_rewrites list is traversed in multiple places. The overhead
should be negligible in comparison to a rewrite and the list should not contain
a lot of rewrites in practice.

The pending_rewrites list is traversed in multiple places. In typical usage,
the list will contain only a few entries so the traversal cost is negligible (
furthermore in comparison to a rewrite).
---
 src/backend/catalog/index.c                  |   2 +-
 src/backend/commands/cluster.c               |   5 +
 src/backend/commands/tablecmds.c             |   6 +
 src/backend/utils/activity/pgstat_relation.c | 391 ++++++++++++++++++-
 src/backend/utils/activity/pgstat_xact.c     |  25 +-
 src/backend/utils/cache/relcache.c           |   6 +
 src/include/pgstat.h                         |   5 +-
 src/tools/pgindent/typedefs.list             |   1 +
 8 files changed, 424 insertions(+), 17 deletions(-)
  92.8% src/backend/utils/activity/
   4.9% src/backend/

diff --git a/src/backend/catalog/index.c b/src/backend/catalog/index.c
index 5d9db167e59..8b6a7652fcf 100644
--- a/src/backend/catalog/index.c
+++ b/src/backend/catalog/index.c
@@ -1795,7 +1795,7 @@ index_concurrently_swap(Oid newIndexId, Oid oldIndexId, const char *oldName)
 	changeDependenciesOn(RelationRelationId, oldIndexId, newIndexId);
 
 	/* copy over statistics from old to new index */
-	pgstat_copy_relation_stats(newClassRel, oldClassRel);
+	pgstat_copy_relation_stats(newClassRel->rd_locator, oldClassRel->rd_locator, false);
 
 	/* Copy data of pg_statistic from the old index to the new one */
 	CopyStatistics(oldIndexId, newIndexId);
diff --git a/src/backend/commands/cluster.c b/src/backend/commands/cluster.c
index b55221d44cd..da75dfa6ab8 100644
--- a/src/backend/commands/cluster.c
+++ b/src/backend/commands/cluster.c
@@ -1196,6 +1196,11 @@ swap_relation_files(Oid r1, Oid r2, bool target_is_pg_class,
 
 		rel1 = relation_open(r1, NoLock);
 		rel2 = relation_open(r2, NoLock);
+
+		/* Mark that a rewrite happened */
+		if (RELKIND_HAS_STORAGE(rel1->rd_rel->relkind))
+			pgstat_mark_rewrite(rel1->rd_locator, rel2->rd_locator);
+
 		rel2->rd_createSubid = rel1->rd_createSubid;
 		rel2->rd_newRelfilelocatorSubid = rel1->rd_newRelfilelocatorSubid;
 		rel2->rd_firstRelfilelocatorSubid = rel1->rd_firstRelfilelocatorSubid;
diff --git a/src/backend/commands/tablecmds.c b/src/backend/commands/tablecmds.c
index 3aac459e483..540923452fb 100644
--- a/src/backend/commands/tablecmds.c
+++ b/src/backend/commands/tablecmds.c
@@ -16848,6 +16848,7 @@ ATExecSetTableSpace(Oid tableOid, Oid newTableSpace, LOCKMODE lockmode)
 	Oid			reltoastrelid;
 	RelFileNumber newrelfilenumber;
 	RelFileLocator newrlocator;
+	RelFileLocator oldrlocator;
 	List	   *reltoastidxids = NIL;
 	ListCell   *lc;
 
@@ -16886,6 +16887,7 @@ ATExecSetTableSpace(Oid tableOid, Oid newTableSpace, LOCKMODE lockmode)
 	newrlocator = rel->rd_locator;
 	newrlocator.relNumber = newrelfilenumber;
 	newrlocator.spcOid = newTableSpace;
+	oldrlocator = rel->rd_locator;
 
 	/* hand off to AM to actually create new rel storage and copy the data */
 	if (rel->rd_rel->relkind == RELKIND_INDEX)
@@ -16898,6 +16900,10 @@ ATExecSetTableSpace(Oid tableOid, Oid newTableSpace, LOCKMODE lockmode)
 		table_relation_copy_data(rel, &newrlocator);
 	}
 
+	/* mark that a rewrite happened */
+	if (RELKIND_HAS_STORAGE(rel->rd_rel->relkind))
+		pgstat_mark_rewrite(oldrlocator, newrlocator);
+
 	/*
 	 * Update the pg_class row.
 	 *
diff --git a/src/backend/utils/activity/pgstat_relation.c b/src/backend/utils/activity/pgstat_relation.c
index 7debb14bb5d..15b4663eb77 100644
--- a/src/backend/utils/activity/pgstat_relation.c
+++ b/src/backend/utils/activity/pgstat_relation.c
@@ -30,6 +30,19 @@
 #include "utils/syscache.h"
 #include "utils/timestamp.h"
 
+/* Pending rewrite operations for stats copying */
+typedef struct PgStat_PendingRewrite
+{
+	RelFileLocator old_locator;
+	RelFileLocator new_locator;
+	RelFileLocator original_locator;
+	int			nest_level;		/* Transaction nesting level where rewrite
+								 * occurred */
+	struct PgStat_PendingRewrite *next;
+} PgStat_PendingRewrite;
+
+/* The pending rewrites list for current transaction */
+static PgStat_PendingRewrite *pending_rewrites = NULL;
 
 /* Record that's written to 2PC state file when pgstat state is persisted */
 typedef struct TwoPhasePgStatRecord
@@ -43,6 +56,8 @@ typedef struct TwoPhasePgStatRecord
 	PgStat_Counter deleted_pre_truncdrop;
 	RelFileLocator locator;		/* table's rd_locator */
 	bool		truncdropped;	/* was the relation truncated/dropped? */
+	RelFileLocator rewrite_old_locator;
+	int			rewrite_nest_level;
 } TwoPhasePgStatRecord;
 
 
@@ -54,27 +69,70 @@ static void restore_truncdrop_counters(PgStat_TableXactStatus *trans);
 
 
 /*
- * Copy stats between relations. This is used for things like REINDEX
+ * Copy stats between RelFileLocator. This is used for things like REINDEX
  * CONCURRENTLY.
  */
 void
-pgstat_copy_relation_stats(Relation dst, Relation src)
+pgstat_copy_relation_stats(RelFileLocator dst, RelFileLocator src, bool increment)
 {
 	PgStat_StatTabEntry *srcstats;
 	PgStatShared_Relation *dstshstats;
 	PgStat_EntryRef *dst_ref;
 
-	srcstats = pgstat_fetch_stat_tabentry_ext(RelationGetRelid(src));
+	srcstats = (PgStat_StatTabEntry *) pgstat_fetch_entry(PGSTAT_KIND_RELATION,
+														  src.dbOid,
+														  RelFileLocatorToPgStatObjid(src));
 	if (!srcstats)
 		return;
 
 	dst_ref = pgstat_get_entry_ref_locked(PGSTAT_KIND_RELATION,
-										  dst->rd_rel->relisshared ? InvalidOid : MyDatabaseId,
-										  RelationGetRelid(dst),
+										  dst.dbOid,
+										  RelFileLocatorToPgStatObjid(dst),
 										  false);
 
 	dstshstats = (PgStatShared_Relation *) dst_ref->shared_stats;
-	dstshstats->stats = *srcstats;
+
+	if (!increment)
+		dstshstats->stats = *srcstats;
+	else
+	{
+		/* Increment those statistics */
+#define RELFSTAT_ACC(fld, stats_to_add) \
+	(dstshstats->stats.fld += stats_to_add->fld)
+		RELFSTAT_ACC(numscans, srcstats);
+		RELFSTAT_ACC(tuples_returned, srcstats);
+		RELFSTAT_ACC(tuples_fetched, srcstats);
+		RELFSTAT_ACC(tuples_inserted, srcstats);
+		RELFSTAT_ACC(tuples_updated, srcstats);
+		RELFSTAT_ACC(tuples_deleted, srcstats);
+		RELFSTAT_ACC(tuples_hot_updated, srcstats);
+		RELFSTAT_ACC(tuples_newpage_updated, srcstats);
+		RELFSTAT_ACC(live_tuples, srcstats);
+		RELFSTAT_ACC(dead_tuples, srcstats);
+		RELFSTAT_ACC(mod_since_analyze, srcstats);
+		RELFSTAT_ACC(ins_since_vacuum, srcstats);
+		RELFSTAT_ACC(blocks_fetched, srcstats);
+		RELFSTAT_ACC(blocks_hit, srcstats);
+		RELFSTAT_ACC(vacuum_count, srcstats);
+		RELFSTAT_ACC(autovacuum_count, srcstats);
+		RELFSTAT_ACC(analyze_count, srcstats);
+		RELFSTAT_ACC(autoanalyze_count, srcstats);
+		RELFSTAT_ACC(total_vacuum_time, srcstats);
+		RELFSTAT_ACC(total_autovacuum_time, srcstats);
+		RELFSTAT_ACC(total_analyze_time, srcstats);
+		RELFSTAT_ACC(total_autoanalyze_time, srcstats);
+#undef RELFSTAT_ACC
+
+		/* Replace those statistics */
+#define RELFSTAT_REP(fld, stats_to_rep) \
+	(dstshstats->stats.fld = stats_to_rep->fld)
+		RELFSTAT_REP(lastscan, srcstats);
+		RELFSTAT_REP(last_vacuum_time, srcstats);
+		RELFSTAT_REP(last_autovacuum_time, srcstats);
+		RELFSTAT_REP(last_analyze_time, srcstats);
+		RELFSTAT_REP(last_autoanalyze_time, srcstats);
+#undef RELFSTAT_REP
+	}
 
 	pgstat_unlock_entry(dst_ref);
 }
@@ -136,6 +194,7 @@ void
 pgstat_assoc_relation(Relation rel)
 {
 	RelFileLocator locator;
+	PgStat_TableStatus *pgstat_info;
 
 	Assert(rel->pgstat_enabled);
 	Assert(rel->pgstat_info == NULL);
@@ -165,14 +224,54 @@ pgstat_assoc_relation(Relation rel)
 		locator.relNumber = rel->rd_id;
 	}
 
+	/*
+	 * If this relation was rewritten during the current transaction we may be
+	 * reopening it with its new RelFileLocator. In that case, continue using
+	 * the stats entry associated with the old locator rather than creating a
+	 * new one. This ensures all stats from before and after the rewrite are
+	 * tracked in a single entry which will be properly copied to the new
+	 * locator at transaction commit.
+	 */
+	if (pending_rewrites != NULL)
+	{
+		PgStat_PendingRewrite *rewrite;
+
+		for (rewrite = pending_rewrites; rewrite != NULL; rewrite = rewrite->next)
+		{
+			if (locator.dbOid == rewrite->new_locator.dbOid &&
+				locator.spcOid == rewrite->new_locator.spcOid &&
+				locator.relNumber == rewrite->new_locator.relNumber)
+			{
+				pgstat_info = pgstat_prep_relation_pending(rewrite->old_locator);
+				goto found_entry;
+			}
+		}
+	}
+
 	/* Else find or make the PgStat_TableStatus entry, and update link */
-	rel->pgstat_info = pgstat_prep_relation_pending(locator);
+	pgstat_info = pgstat_prep_relation_pending(locator);
+
+found_entry:
+	rel->pgstat_info = pgstat_info;
+
+	/*
+	 * For relations stats, we key by physical file location, not by relation
+	 * OID. This means during operations like ALTER TYPE it's possible that
+	 * the relation OID changes but the relfilenode stays the same (no actual
+	 * rewrite needed). Unlink the old relation first.
+	 */
+	if (pgstat_info->relation != NULL &&
+		pgstat_info->relation != rel)
+	{
+		pgstat_info->relation->pgstat_info = NULL;
+		pgstat_info->relation = NULL;
+	}
 
 	/* don't allow link a stats to multiple relcache entries */
-	Assert(rel->pgstat_info->relation == NULL);
+	Assert(pgstat_info->relation == NULL);
 
 	/* mark this relation as the owner */
-	rel->pgstat_info->relation = rel;
+	pgstat_info->relation = rel;
 }
 
 /*
@@ -215,14 +314,37 @@ pgstat_drop_relation(Relation rel)
 {
 	int			nest_level = GetCurrentTransactionNestLevel();
 	PgStat_TableStatus *pgstat_info;
+	bool		skip_transactional_drop = false;
 
 	/* don't track stats for relations without storage */
 	if (!RELKIND_HAS_STORAGE(rel->rd_rel->relkind))
 		return;
 
-	pgstat_drop_transactional(PGSTAT_KIND_RELATION,
-							  rel->rd_locator.dbOid,
-							  RelFileLocatorToPgStatObjid(rel->rd_locator));
+	/* Check if this drop is part of a pending rewrite */
+	if (pending_rewrites != NULL)
+	{
+		PgStat_PendingRewrite *rewrite;
+
+		for (rewrite = pending_rewrites; rewrite != NULL; rewrite = rewrite->next)
+		{
+			if (rel->rd_locator.dbOid == rewrite->old_locator.dbOid &&
+				rel->rd_locator.spcOid == rewrite->old_locator.spcOid &&
+				rel->rd_locator.relNumber == rewrite->old_locator.relNumber)
+			{
+				skip_transactional_drop = true;
+				break;
+			}
+		}
+	}
+
+	/*
+	 * If it is part of a rewrite, drop its stats later, for example in
+	 * AtEOXact_PgStat_Relations(), so skip it here.
+	 */
+	if (!skip_transactional_drop)
+		pgstat_drop_transactional(PGSTAT_KIND_RELATION,
+								  rel->rd_locator.dbOid,
+								  RelFileLocatorToPgStatObjid(rel->rd_locator));
 
 	if (!pgstat_should_count_relation(rel))
 		return;
@@ -660,6 +782,48 @@ AtEOXact_PgStat_Relations(PgStat_SubXactStatus *xact_state, bool isCommit)
 		}
 		tabstat->trans = NULL;
 	}
+
+	/* preserve the stats in case of rewrite */
+	if (isCommit && pending_rewrites != NULL)
+	{
+		PgStat_PendingRewrite *rewrite;
+		PgStat_PendingRewrite *prev = NULL;
+		PgStat_PendingRewrite *current = pending_rewrites;
+		PgStat_PendingRewrite *next;
+
+		/* reverse the rewrites list to process in chronological order */
+		while (current != NULL)
+		{
+			next = current->next;
+			current->next = prev;
+			prev = current;
+			current = next;
+		}
+
+		/* now process rewrites in chronological order */
+		for (rewrite = prev; rewrite != NULL; rewrite = rewrite->next)
+		{
+			PgStat_EntryRef *old_entry_ref;
+
+			old_entry_ref = pgstat_fetch_pending_entry(PGSTAT_KIND_RELATION,
+													   rewrite->old_locator.dbOid,
+													   RelFileLocatorToPgStatObjid(rewrite->old_locator));
+
+			if (old_entry_ref && old_entry_ref->pending)
+				pgstat_relation_flush_cb(old_entry_ref, false);
+
+			pgstat_copy_relation_stats(rewrite->new_locator,
+									   rewrite->old_locator, true);
+
+			/* drop old locator's stats */
+			if (!pgstat_drop_entry(PGSTAT_KIND_RELATION,
+								   rewrite->old_locator.dbOid,
+								   RelFileLocatorToPgStatObjid(rewrite->old_locator)))
+				pgstat_request_entry_refs_gc();
+		}
+	}
+
+	pending_rewrites = NULL;
 }
 
 /*
@@ -675,6 +839,30 @@ AtEOSubXact_PgStat_Relations(PgStat_SubXactStatus *xact_state, bool isCommit, in
 	PgStat_TableXactStatus *trans;
 	PgStat_TableXactStatus *next_trans;
 
+	/*
+	 * If we don't commit then remove the associated rewrites if any, to keep
+	 * the rewrite chain in sync with what will be eventually committed.
+	 */
+	if (!isCommit)
+	{
+		PgStat_PendingRewrite **rewrite_ptr = &pending_rewrites;
+
+		while (*rewrite_ptr != NULL)
+		{
+			if ((*rewrite_ptr)->nest_level >= nestDepth)
+			{
+				PgStat_PendingRewrite *to_remove = *rewrite_ptr;
+
+				*rewrite_ptr = (*rewrite_ptr)->next;
+				pfree(to_remove);
+			}
+			else
+			{
+				rewrite_ptr = &((*rewrite_ptr)->next);
+			}
+		}
+	}
+
 	for (trans = xact_state->first; trans != NULL; trans = next_trans)
 	{
 		PgStat_TableStatus *tabstat;
@@ -754,11 +942,19 @@ void
 AtPrepare_PgStat_Relations(PgStat_SubXactStatus *xact_state)
 {
 	PgStat_TableXactStatus *trans;
+	PgStat_PendingRewrite *rewrite;
 
+	/*
+	 * For each tabstat, find its matching rewrite and remove it from the
+	 * pending rewrites list. This way, after processing all tabstats, pending
+	 * rewrites will only contain rewrite only transactions.
+	 */
 	for (trans = xact_state->first; trans != NULL; trans = trans->next)
 	{
 		PgStat_TableStatus *tabstat PG_USED_FOR_ASSERTS_ONLY;
 		TwoPhasePgStatRecord record;
+		PgStat_PendingRewrite **rewrite_ptr;
+		bool		found_rewrite = false;
 
 		Assert(trans->nest_level == 1);
 		Assert(trans->upper == NULL);
@@ -778,10 +974,83 @@ AtPrepare_PgStat_Relations(PgStat_SubXactStatus *xact_state)
 			record.locator = tabstat->locator;
 
 		record.truncdropped = trans->truncdropped;
+		record.rewrite_nest_level = 0;
+
+		/*
+		 * Look for a matching rewrite and remove it from pending rewrites. We
+		 * check three possible matches:
+		 *
+		 * The new_locator when stats have been added after the rewrite. The
+		 * old_locator when stats have been added before the rewrite but not
+		 * after. The original_locator when this tabstat is part of a rewrite
+		 * chain.
+		 */
+		rewrite_ptr = &pending_rewrites;
+		while (*rewrite_ptr != NULL)
+		{
+			rewrite = *rewrite_ptr;
+
+			if ((record.locator.dbOid == rewrite->new_locator.dbOid &&
+				 record.locator.spcOid == rewrite->new_locator.spcOid &&
+				 record.locator.relNumber == rewrite->new_locator.relNumber) ||
+				(tabstat->locator.dbOid == rewrite->old_locator.dbOid &&
+				 tabstat->locator.spcOid == rewrite->old_locator.spcOid &&
+				 tabstat->locator.relNumber == rewrite->old_locator.relNumber) ||
+				(tabstat->locator.dbOid == rewrite->original_locator.dbOid &&
+				 tabstat->locator.spcOid == rewrite->original_locator.spcOid &&
+				 tabstat->locator.relNumber == rewrite->original_locator.relNumber))
+			{
+				/*
+				 * Found matching rewrite. Record the rewrite information and
+				 * remove this rewrite from the list since it's now handled.
+				 */
+				record.rewrite_old_locator = rewrite->original_locator;
+				record.rewrite_nest_level = rewrite->nest_level;
+				record.locator = rewrite->new_locator;
+				found_rewrite = true;
+
+				/* Remove from pending_rewrites list */
+				*rewrite_ptr = rewrite->next;
+				pfree(rewrite);
+				break;
+			}
+			else
+			{
+				/* Move to next rewrite in the list */
+				rewrite_ptr = &(rewrite->next);
+			}
+		}
+
+		/* If no rewrite found, clear the rewrite fields */
+		if (!found_rewrite)
+		{
+			memset(&record.rewrite_old_locator, 0, sizeof(RelFileLocator));
+		}
+
+		RegisterTwoPhaseRecord(TWOPHASE_RM_PGSTAT_ID, 0,
+							   &record, sizeof(TwoPhasePgStatRecord));
+	}
+
+	/*
+	 * Now process any rewrites still pending. These are rewrite only
+	 * transactions. We need to preserve their stats even though there's no
+	 * tabstat entry for them.
+	 */
+	for (rewrite = pending_rewrites; rewrite != NULL; rewrite = rewrite->next)
+	{
+		TwoPhasePgStatRecord record;
+
+		memset(&record, 0, sizeof(TwoPhasePgStatRecord));
+		record.locator = rewrite->new_locator;
+		record.rewrite_old_locator = rewrite->original_locator;
+		record.rewrite_nest_level = rewrite->nest_level;
+		record.truncdropped = false;
 
 		RegisterTwoPhaseRecord(TWOPHASE_RM_PGSTAT_ID, 0,
 							   &record, sizeof(TwoPhasePgStatRecord));
 	}
+
+	pending_rewrites = NULL;
 }
 
 /*
@@ -804,6 +1073,8 @@ PostPrepare_PgStat_Relations(PgStat_SubXactStatus *xact_state)
 		tabstat = trans->parent;
 		tabstat->trans = NULL;
 	}
+
+	pending_rewrites = NULL;
 }
 
 /*
@@ -839,6 +1110,29 @@ pgstat_twophase_postcommit(FullTransactionId fxid, uint16 info,
 	pgstat_info->counts.changed_tuples +=
 		rec->tuples_inserted + rec->tuples_updated +
 		rec->tuples_deleted;
+
+	if (rec->rewrite_nest_level > 0)
+	{
+		PgStat_EntryRef *old_entry_ref;
+
+		/* Flush any pending stats for old locator first */
+		old_entry_ref = pgstat_fetch_pending_entry(PGSTAT_KIND_RELATION,
+												   rec->rewrite_old_locator.dbOid,
+												   RelFileLocatorToPgStatObjid(rec->rewrite_old_locator));
+
+		if (old_entry_ref && old_entry_ref->pending)
+			pgstat_relation_flush_cb(old_entry_ref, false);
+
+		/* Copy stats from old to new locator */
+		pgstat_copy_relation_stats(rec->locator, rec->rewrite_old_locator,
+								   true);
+
+		/* Drop old locator's stats */
+		if (!pgstat_drop_entry(PGSTAT_KIND_RELATION,
+							   rec->rewrite_old_locator.dbOid,
+							   RelFileLocatorToPgStatObjid(rec->rewrite_old_locator)))
+			pgstat_request_entry_refs_gc();
+	}
 }
 
 /*
@@ -853,9 +1147,26 @@ pgstat_twophase_postabort(FullTransactionId fxid, uint16 info,
 {
 	TwoPhasePgStatRecord *rec = (TwoPhasePgStatRecord *) recdata;
 	PgStat_TableStatus *pgstat_info;
+	RelFileLocator target_locator;
+
+	/*
+	 * For aborted transactions with rewrites (like TRUNCATE), we need to
+	 * restore stats to the old locator, not the new one. The new locator
+	 * should be dropped since the rewrite is being rolled back.
+	 */
+	if (rec->rewrite_nest_level > 0)
+	{
+		/* Use the old locator */
+		target_locator = rec->rewrite_old_locator;
+	}
+	else
+	{
+		/* No rewrite, use the original locator */
+		target_locator = rec->locator;
+	}
 
 	/* Find or create a tabstat entry for the target locator */
-	pgstat_info = pgstat_prep_relation_pending(rec->locator);
+	pgstat_info = pgstat_prep_relation_pending(target_locator);
 
 	/* Same math as in AtEOXact_PgStat, abort case */
 	if (rec->truncdropped)
@@ -910,7 +1221,17 @@ pgstat_relation_flush_cb(PgStat_EntryRef *entry_ref, bool nowait)
 	tabentry->numscans += lstats->counts.numscans;
 	if (lstats->counts.numscans)
 	{
-		TimestampTz t = GetCurrentTransactionStopTimestamp();
+		TimestampTz t;
+
+		/*
+		 * Checking the transaction state due to the flush call in
+		 * pgstat_twophase_postcommit() that would break the assertion on the
+		 * state in GetCurrentTransactionStopTimestamp().
+		 */
+		if (!IsTransactionState())
+			t = GetCurrentTransactionStopTimestamp();
+		else
+			t = GetCurrentTimestamp();
 
 		if (t > tabentry->lastscan)
 			tabentry->lastscan = t;
@@ -1162,3 +1483,45 @@ pgstat_reloid_to_relfilelocator(Oid reloid, RelFileLocator *locator)
 	ReleaseSysCache(tuple);
 	return result;
 }
+
+/*
+ * Mark that a relation rewrite has occurred, preserving the original locator
+ * so stats can be copied at transaction commit.
+ */
+void
+pgstat_mark_rewrite(RelFileLocator old_locator, RelFileLocator new_locator)
+{
+	PgStat_PendingRewrite *rewrite;
+	PgStat_PendingRewrite *existing;
+	RelFileLocator original_locator = old_locator;
+
+	for (existing = pending_rewrites; existing != NULL; existing = existing->next)
+	{
+		if (old_locator.dbOid == existing->new_locator.dbOid &&
+			old_locator.spcOid == existing->new_locator.spcOid &&
+			old_locator.relNumber == existing->new_locator.relNumber)
+		{
+			original_locator = existing->original_locator;
+			break;
+		}
+	}
+
+	/* Allocate in TopTransactionContext memory context */
+	rewrite = MemoryContextAlloc(TopTransactionContext,
+								 sizeof(PgStat_PendingRewrite));
+
+	rewrite->old_locator = old_locator;
+	rewrite->new_locator = new_locator;
+	rewrite->original_locator = original_locator;
+	rewrite->nest_level = GetCurrentTransactionNestLevel();
+
+	/* Add to the list */
+	rewrite->next = pending_rewrites;
+	pending_rewrites = rewrite;
+}
+
+void
+pgstat_clear_rewrite(void)
+{
+	pending_rewrites = NULL;
+}
diff --git a/src/backend/utils/activity/pgstat_xact.c b/src/backend/utils/activity/pgstat_xact.c
index bc9864bd8d9..f8cf3755ce2 100644
--- a/src/backend/utils/activity/pgstat_xact.c
+++ b/src/backend/utils/activity/pgstat_xact.c
@@ -55,6 +55,8 @@ AtEOXact_PgStat(bool isCommit, bool parallel)
 	}
 	pgStatXactStack = NULL;
 
+	pgstat_clear_rewrite();
+
 	/* Make sure any stats snapshot is thrown away */
 	pgstat_clear_snapshot();
 }
@@ -360,8 +362,29 @@ create_drop_transactional_internal(PgStat_Kind kind, Oid dboid, uint64 objid, bo
 void
 pgstat_create_transactional(PgStat_Kind kind, Oid dboid, uint64 objid)
 {
-	if (pgstat_get_entry_ref(kind, dboid, objid, false, NULL))
+	PgStat_EntryRef *entry_ref;
+
+	entry_ref = pgstat_get_entry_ref(kind, dboid, objid, false, NULL);
+
+	if (entry_ref)
 	{
+		/*
+		 * For relations stats, we key by physical file location, not by
+		 * relation OID. This means during operations like ALTER TYPE where
+		 * the relation OID changes but the relfilenode stays the same (no
+		 * actual rewrite needed), we'll find an existing entry.
+		 *
+		 * This is expected behavior, we want to preserve stats across the
+		 * catalog change. Simply reset and recreate the entry for the new
+		 * relation OID without warning.
+		 */
+		if (kind == PGSTAT_KIND_RELATION)
+		{
+			pgstat_reset(kind, dboid, objid);
+			create_drop_transactional_internal(kind, dboid, objid, true);
+			return;
+		}
+
 		ereport(WARNING,
 				errmsg("resetting existing statistics for kind %s, db=%u, oid=%" PRIu64,
 					   (pgstat_get_kind_info(kind))->name, dboid,
diff --git a/src/backend/utils/cache/relcache.c b/src/backend/utils/cache/relcache.c
index 915d0bc9084..7a7f8023eb3 100644
--- a/src/backend/utils/cache/relcache.c
+++ b/src/backend/utils/cache/relcache.c
@@ -85,6 +85,7 @@
 #include "utils/inval.h"
 #include "utils/lsyscache.h"
 #include "utils/memutils.h"
+#include "utils/pgstat_internal.h"
 #include "utils/relmapper.h"
 #include "utils/resowner.h"
 #include "utils/snapmgr.h"
@@ -3780,6 +3781,7 @@ RelationSetNewRelfilenumber(Relation relation, char persistence)
 	MultiXactId minmulti = InvalidMultiXactId;
 	TransactionId freezeXid = InvalidTransactionId;
 	RelFileLocator newrlocator;
+	RelFileLocator oldrlocator = relation->rd_locator;
 
 	if (!IsBinaryUpgrade)
 	{
@@ -3951,6 +3953,10 @@ RelationSetNewRelfilenumber(Relation relation, char persistence)
 
 	table_close(pg_class, RowExclusiveLock);
 
+	/* Mark that a rewrite happened */
+	if (RELKIND_HAS_STORAGE(relation->rd_rel->relkind))
+		pgstat_mark_rewrite(oldrlocator, newrlocator);
+
 	/*
 	 * Make the pg_class row change or relation map change visible.  This will
 	 * cause the relcache entry to get updated, too.
diff --git a/src/include/pgstat.h b/src/include/pgstat.h
index 5d0fe79f7e3..332dffde400 100644
--- a/src/include/pgstat.h
+++ b/src/include/pgstat.h
@@ -665,7 +665,7 @@ extern PgStat_FunctionCounts *find_funcstat_entry(Oid func_id);
 
 extern void pgstat_create_relation(Relation rel);
 extern void pgstat_drop_relation(Relation rel);
-extern void pgstat_copy_relation_stats(Relation dst, Relation src);
+extern void pgstat_copy_relation_stats(RelFileLocator dst, RelFileLocator src, bool increment);
 
 extern void pgstat_init_relation(Relation rel);
 extern void pgstat_assoc_relation(Relation rel);
@@ -677,6 +677,9 @@ extern void pgstat_report_vacuum(RelFileLocator locator, PgStat_Counter livetupl
 extern void pgstat_report_analyze(Relation rel,
 								  PgStat_Counter livetuples, PgStat_Counter deadtuples,
 								  bool resetcounter, TimestampTz starttime);
+extern void pgstat_mark_rewrite(RelFileLocator old_locator,
+								RelFileLocator new_locator);
+extern void pgstat_clear_rewrite(void);
 
 /*
  * If stats are enabled, but pending data hasn't been prepared yet, call
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 432509277c9..bd8fcb16dcf 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -2248,6 +2248,7 @@ PgStat_KindInfo
 PgStat_LocalState
 PgStat_PendingDroppedStatsItem
 PgStat_PendingIO
+PgStat_PendingRewrite
 PgStat_SLRUStats
 PgStat_ShmemControl
 PgStat_Snapshot
-- 
2.34.1


--KMYjwlsACqdYx2P8--





^ permalink  raw  reply  [nested|flat] 24+ messages in thread


end of thread, other threads:[~2025-11-04 13:52 UTC | newest]

Thread overview: 24+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2022-01-05 19:24 [PATCH v7 1/6] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v18 1/4] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v14 1/3] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v12 1/6] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v13 1/6] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v16 1/4] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v15 1/4] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v17 1/4] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v19 1/4] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v20 1/4] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v5 1/8] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v12 1/6] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v14 1/3] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v4 1/8] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v17 1/4] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v11 1/6] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v9 1/6] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v15 1/4] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v10 1/6] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v13 1/6] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v6 1/6] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v16 1/4] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2022-01-05 19:24 [PATCH v8 1/6] Introduce custodian. Nathan Bossart <bossartn@amazon.com>
2025-11-04 13:52 [PATCH v7 3/3] handle relation statistics correctly during rewrites Bertrand Drouvot <bertranddrouvot.pg@gmail.com>

This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox