pg.ddx.io pgsql-hackers@postgresql.org mailing list archive
help / color / mirror / Atom feednotice processors for isolationtester
4+ messages / 3 participants
[nested] [flat]
Commits from this thread:
2018-11-09 a28e10e82e Indicate session name in isolationtester notices
* notice processors for isolationtester
@ 2018-10-24 21:34 Alvaro Herrera <alvherre@2ndquadrant.com>
0 siblings, 2 replies; 4+ messages in thread
From: Alvaro Herrera @ 2018-10-24 21:34 UTC (permalink / raw)
To: pgsql-hackers
Recently while testing a patch I found it immensely useful to
distinguish which session each WARNING message came from, when bespoke
tests were run under isolationtester. Current code does not show that,
so I developed this patch adding notice processors, and then it does.
--
Álvaro Herrera https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
Attachments:
isolationtester-notices.patch (text/x-diff, 9.6K) download
commit dbbd8840ab070ced5dcd1eefaa67319bd9747b0c[m
Author: Alvaro Herrera <alvherre@alvh.no-ip.org>
AuthorDate: Thu Sep 27 16:37:43 2018 -0300
CommitDate: Wed Oct 24 18:31:00 2018 -0300
Have isolationtester prefix WARNING with session name
diff --git a/src/test/isolation/expected/vacuum-concurrent-drop.out b/src/test/isolation/expected/vacuum-concurrent-drop.out
index bb92e80b92..cf348d7e5d 100644
--- a/src/test/isolation/expected/vacuum-concurrent-drop.out
+++ b/src/test/isolation/expected/vacuum-concurrent-drop.out
@@ -10,7 +10,7 @@ step drop_and_commit:
DROP TABLE part2;
COMMIT;
-WARNING: skipping vacuum of "part2" --- relation no longer exists
+s2: WARNING: skipping vacuum of "part2" --- relation no longer exists
step vac_specified: <... completed>
starting permutation: lock vac_all_parts drop_and_commit
@@ -35,7 +35,7 @@ step drop_and_commit:
DROP TABLE part2;
COMMIT;
-WARNING: skipping analyze of "part2" --- relation no longer exists
+s2: WARNING: skipping analyze of "part2" --- relation no longer exists
step analyze_specified: <... completed>
starting permutation: lock analyze_all_parts drop_and_commit
@@ -60,7 +60,7 @@ step drop_and_commit:
DROP TABLE part2;
COMMIT;
-WARNING: skipping vacuum of "part2" --- relation no longer exists
+s2: WARNING: skipping vacuum of "part2" --- relation no longer exists
step vac_analyze_specified: <... completed>
starting permutation: lock vac_analyze_all_parts drop_and_commit
diff --git a/src/test/isolation/expected/vacuum-conflict.out b/src/test/isolation/expected/vacuum-conflict.out
index 06ac75ef23..ffde537305 100644
--- a/src/test/isolation/expected/vacuum-conflict.out
+++ b/src/test/isolation/expected/vacuum-conflict.out
@@ -4,7 +4,7 @@ starting permutation: s1_begin s1_lock s2_auth s2_vacuum s1_commit s2_reset
step s1_begin: BEGIN;
step s1_lock: LOCK vacuum_tab IN SHARE UPDATE EXCLUSIVE MODE;
step s2_auth: SET ROLE regress_vacuum_conflict;
-WARNING: skipping "vacuum_tab" --- only table or database owner can vacuum it
+s2: WARNING: skipping "vacuum_tab" --- only table or database owner can vacuum it
step s2_vacuum: VACUUM vacuum_tab;
step s1_commit: COMMIT;
step s2_reset: RESET ROLE;
@@ -12,7 +12,7 @@ step s2_reset: RESET ROLE;
starting permutation: s1_begin s2_auth s2_vacuum s1_lock s1_commit s2_reset
step s1_begin: BEGIN;
step s2_auth: SET ROLE regress_vacuum_conflict;
-WARNING: skipping "vacuum_tab" --- only table or database owner can vacuum it
+s2: WARNING: skipping "vacuum_tab" --- only table or database owner can vacuum it
step s2_vacuum: VACUUM vacuum_tab;
step s1_lock: LOCK vacuum_tab IN SHARE UPDATE EXCLUSIVE MODE;
step s1_commit: COMMIT;
@@ -22,14 +22,14 @@ starting permutation: s1_begin s2_auth s1_lock s2_vacuum s1_commit s2_reset
step s1_begin: BEGIN;
step s2_auth: SET ROLE regress_vacuum_conflict;
step s1_lock: LOCK vacuum_tab IN SHARE UPDATE EXCLUSIVE MODE;
-WARNING: skipping "vacuum_tab" --- only table or database owner can vacuum it
+s2: WARNING: skipping "vacuum_tab" --- only table or database owner can vacuum it
step s2_vacuum: VACUUM vacuum_tab;
step s1_commit: COMMIT;
step s2_reset: RESET ROLE;
starting permutation: s2_auth s2_vacuum s1_begin s1_lock s1_commit s2_reset
step s2_auth: SET ROLE regress_vacuum_conflict;
-WARNING: skipping "vacuum_tab" --- only table or database owner can vacuum it
+s2: WARNING: skipping "vacuum_tab" --- only table or database owner can vacuum it
step s2_vacuum: VACUUM vacuum_tab;
step s1_begin: BEGIN;
step s1_lock: LOCK vacuum_tab IN SHARE UPDATE EXCLUSIVE MODE;
@@ -40,7 +40,7 @@ starting permutation: s1_begin s1_lock s2_auth s2_analyze s1_commit s2_reset
step s1_begin: BEGIN;
step s1_lock: LOCK vacuum_tab IN SHARE UPDATE EXCLUSIVE MODE;
step s2_auth: SET ROLE regress_vacuum_conflict;
-WARNING: skipping "vacuum_tab" --- only table or database owner can analyze it
+s2: WARNING: skipping "vacuum_tab" --- only table or database owner can analyze it
step s2_analyze: ANALYZE vacuum_tab;
step s1_commit: COMMIT;
step s2_reset: RESET ROLE;
@@ -48,7 +48,7 @@ step s2_reset: RESET ROLE;
starting permutation: s1_begin s2_auth s2_analyze s1_lock s1_commit s2_reset
step s1_begin: BEGIN;
step s2_auth: SET ROLE regress_vacuum_conflict;
-WARNING: skipping "vacuum_tab" --- only table or database owner can analyze it
+s2: WARNING: skipping "vacuum_tab" --- only table or database owner can analyze it
step s2_analyze: ANALYZE vacuum_tab;
step s1_lock: LOCK vacuum_tab IN SHARE UPDATE EXCLUSIVE MODE;
step s1_commit: COMMIT;
@@ -58,14 +58,14 @@ starting permutation: s1_begin s2_auth s1_lock s2_analyze s1_commit s2_reset
step s1_begin: BEGIN;
step s2_auth: SET ROLE regress_vacuum_conflict;
step s1_lock: LOCK vacuum_tab IN SHARE UPDATE EXCLUSIVE MODE;
-WARNING: skipping "vacuum_tab" --- only table or database owner can analyze it
+s2: WARNING: skipping "vacuum_tab" --- only table or database owner can analyze it
step s2_analyze: ANALYZE vacuum_tab;
step s1_commit: COMMIT;
step s2_reset: RESET ROLE;
starting permutation: s2_auth s2_analyze s1_begin s1_lock s1_commit s2_reset
step s2_auth: SET ROLE regress_vacuum_conflict;
-WARNING: skipping "vacuum_tab" --- only table or database owner can analyze it
+s2: WARNING: skipping "vacuum_tab" --- only table or database owner can analyze it
step s2_analyze: ANALYZE vacuum_tab;
step s1_begin: BEGIN;
step s1_lock: LOCK vacuum_tab IN SHARE UPDATE EXCLUSIVE MODE;
diff --git a/src/test/isolation/expected/vacuum-skip-locked.out b/src/test/isolation/expected/vacuum-skip-locked.out
index 95ca4569cc..99db281a15 100644
--- a/src/test/isolation/expected/vacuum-skip-locked.out
+++ b/src/test/isolation/expected/vacuum-skip-locked.out
@@ -5,7 +5,7 @@ step lock_share:
BEGIN;
LOCK part1 IN SHARE MODE;
-WARNING: skipping vacuum of "part1" --- lock not available
+s2: WARNING: skipping vacuum of "part1" --- lock not available
step vac_specified: VACUUM (SKIP_LOCKED) part1, part2;
step commit:
COMMIT;
@@ -26,7 +26,7 @@ step lock_share:
BEGIN;
LOCK part1 IN SHARE MODE;
-WARNING: skipping analyze of "part1" --- lock not available
+s2: WARNING: skipping analyze of "part1" --- lock not available
step analyze_specified: ANALYZE (SKIP_LOCKED) part1, part2;
step commit:
COMMIT;
@@ -47,7 +47,7 @@ step lock_share:
BEGIN;
LOCK part1 IN SHARE MODE;
-WARNING: skipping vacuum of "part1" --- lock not available
+s2: WARNING: skipping vacuum of "part1" --- lock not available
step vac_analyze_specified: VACUUM (ANALYZE, SKIP_LOCKED) part1, part2;
step commit:
COMMIT;
@@ -68,7 +68,7 @@ step lock_share:
BEGIN;
LOCK part1 IN SHARE MODE;
-WARNING: skipping vacuum of "part1" --- lock not available
+s2: WARNING: skipping vacuum of "part1" --- lock not available
step vac_full_specified: VACUUM (SKIP_LOCKED, FULL) part1, part2;
step commit:
COMMIT;
@@ -89,7 +89,7 @@ step lock_access_exclusive:
BEGIN;
LOCK part1 IN ACCESS EXCLUSIVE MODE;
-WARNING: skipping vacuum of "part1" --- lock not available
+s2: WARNING: skipping vacuum of "part1" --- lock not available
step vac_specified: VACUUM (SKIP_LOCKED) part1, part2;
step commit:
COMMIT;
@@ -110,7 +110,7 @@ step lock_access_exclusive:
BEGIN;
LOCK part1 IN ACCESS EXCLUSIVE MODE;
-WARNING: skipping analyze of "part1" --- lock not available
+s2: WARNING: skipping analyze of "part1" --- lock not available
step analyze_specified: ANALYZE (SKIP_LOCKED) part1, part2;
step commit:
COMMIT;
@@ -132,7 +132,7 @@ step lock_access_exclusive:
BEGIN;
LOCK part1 IN ACCESS EXCLUSIVE MODE;
-WARNING: skipping vacuum of "part1" --- lock not available
+s2: WARNING: skipping vacuum of "part1" --- lock not available
step vac_analyze_specified: VACUUM (ANALYZE, SKIP_LOCKED) part1, part2;
step commit:
COMMIT;
@@ -154,7 +154,7 @@ step lock_access_exclusive:
BEGIN;
LOCK part1 IN ACCESS EXCLUSIVE MODE;
-WARNING: skipping vacuum of "part1" --- lock not available
+s2: WARNING: skipping vacuum of "part1" --- lock not available
step vac_full_specified: VACUUM (SKIP_LOCKED, FULL) part1, part2;
step commit:
COMMIT;
diff --git a/src/test/isolation/isolationtester.c b/src/test/isolation/isolationtester.c
index 62097ab961..e2638553f6 100644
--- a/src/test/isolation/isolationtester.c
+++ b/src/test/isolation/isolationtester.c
@@ -48,6 +48,8 @@ static int step_qsort_cmp(const void *a, const void *b);
static int step_bsearch_cmp(const void *a, const void *b);
static void printResultSet(PGresult *res);
+static void isotesterNoticeProcessor(void *arg, const char *message);
+static void blackholeNoticeProcessor(void *arg, const char *message);
/* close all connections and exit */
static void
@@ -172,6 +174,21 @@ main(int argc, char **argv)
}
/*
+ * Set up notice processors for the user-defined connections, so that
+ * messages can get printed prefixed with the session names. The
+ * control connection gets a "blackhole" processor instead (hides all
+ * messages).
+ */
+ if (i != 0)
+ PQsetNoticeProcessor(conns[i],
+ isotesterNoticeProcessor,
+ (void *) (testspec->sessions[i - 1]->name));
+ else
+ PQsetNoticeProcessor(conns[i],
+ blackholeNoticeProcessor,
+ NULL);
+
+ /*
* Suppress NOTIFY messages, which otherwise pop into results at odd
* places.
*/
@@ -881,3 +898,17 @@ printResultSet(PGresult *res)
printf("\n");
}
}
+
+/* notice processor, prefixes each message with the session name */
+static void
+isotesterNoticeProcessor(void *arg, const char *message)
+{
+ fprintf(stderr, "%s: %s", (char *) arg, message);
+}
+
+/* notice processor, hides the message */
+static void
+blackholeNoticeProcessor(void *arg, const char *message)
+{
+ /* do nothing */
+}
* Re: notice processors for isolationtester
@ 2018-10-24 21:38 Andres Freund <andres@anarazel.de>
parent: Alvaro Herrera <alvherre@2ndquadrant.com>
1 sibling, 0 replies; 4+ messages in thread
From: Andres Freund @ 2018-10-24 21:38 UTC (permalink / raw)
To: Alvaro Herrera <alvherre@2ndquadrant.com>; +Cc: pgsql-hackers
On 2018-10-24 18:34:51 -0300, Alvaro Herrera wrote:
> Recently while testing a patch I found it immensely useful to
> distinguish which session each WARNING message came from, when bespoke
> tests were run under isolationtester. Current code does not show that,
> so I developed this patch adding notice processors, and then it does.
Neat!
^ permalink raw reply [nested|flat] 4+ messages in thread
* Re: notice processors for isolationtester
@ 2018-11-09 04:17 Haribabu Kommi <kommi.haribabu@gmail.com>
parent: Alvaro Herrera <alvherre@2ndquadrant.com>
1 sibling, 1 reply; 4+ messages in thread
From: Haribabu Kommi @ 2018-11-09 04:17 UTC (permalink / raw)
To: Alvaro Herrera <alvherre@2ndquadrant.com>; +Cc: pgsql-hackers
On Thu, Oct 25, 2018 at 8:35 AM Alvaro Herrera <alvherre@2ndquadrant.com>
wrote:
> Recently while testing a patch I found it immensely useful to
> distinguish which session each WARNING message came from, when bespoke
> tests were run under isolationtester. Current code does not show that,
> so I developed this patch adding notice processors, and then it does.
>
I reviewed and tested the patch. It is working as expected.
Providing session name helps in debugging.
Marking the patch as ready for committer.
Regards,
Haribabu Kommi
Fujitsu Australia
^ permalink raw reply [nested|flat] 4+ messages in thread
* Re: notice processors for isolationtester
@ 2018-11-09 16:16 Alvaro Herrera <alvherre@2ndquadrant.com>
parent: Haribabu Kommi <kommi.haribabu@gmail.com>
0 siblings, 0 replies; 4+ messages in thread
From: Alvaro Herrera @ 2018-11-09 16:16 UTC (permalink / raw)
To: Haribabu Kommi <kommi.haribabu@gmail.com>; +Cc: pgsql-hackers
On 2018-Nov-09, Haribabu Kommi wrote:
> On Thu, Oct 25, 2018 at 8:35 AM Alvaro Herrera <alvherre@2ndquadrant.com>
> wrote:
>
> > Recently while testing a patch I found it immensely useful to
> > distinguish which session each WARNING message came from, when bespoke
> > tests were run under isolationtester. Current code does not show that,
> > so I developed this patch adding notice processors, and then it does.
>
> I reviewed and tested the patch. It is working as expected.
> Providing session name helps in debugging.
>
> Marking the patch as ready for committer.
Thanks! Pushed.
--
Álvaro Herrera https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
^ permalink raw reply [nested|flat] 4+ messages in thread
end of thread, other threads:[~2018-11-09 16:16 UTC | newest]
Thread overview: 4+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2018-10-24 21:34 notice processors for isolationtester Alvaro Herrera <alvherre@2ndquadrant.com>
2018-10-24 21:38 ` Andres Freund <andres@anarazel.de>
2018-11-09 04:17 ` Haribabu Kommi <kommi.haribabu@gmail.com>
2018-11-09 16:16 ` Alvaro Herrera <alvherre@2ndquadrant.com>
This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox