Horton

Re: A failure in 031_recovery_conflict.pl on Debian/s390x

Christoph Berg <myon@debian.org>

View thread
From: Christoph Berg <myon@debian.org>
To: Thomas Munro <thomas.munro@gmail.com>
Cc: Andres Freund <andres@anarazel.de>, Noah Misch <noah@leadboat.com>, Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2023-08-10T09:15:11Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Fix instability in 031_recovery_conflict.pl.

  2. Fix recovery conflict SIGUSR1 handling.

  3. Disable 031_recovery_conflict.pl in 15 and 16.

Attachments

Re: To Thomas Munro
> 603 iterations later it hit again, but didn't log anything. (I believe
> I did run "make" in the right directory.)

This time it took 3086 iterations to hit the problem.

Running c27f8621eedf7 + Debian patches + v8 +
pgstat-report-conflicts-immediately.patch + the XXX logging.

> [22:20:24.714](3.145s) # issuing query via background psql:
> #     BEGIN;
> #     DECLARE test_recovery_conflict_cursor CURSOR FOR SELECT b FROM test_recovery_conflict_table1;
> #     FETCH FORWARD FROM test_recovery_conflict_cursor;
> [22:20:24.745](0.031s) ok 1 - buffer pin conflict: cursor with conflicting pin established
> Waiting for replication conn standby's replay_lsn to pass 0/3430000 on primary
> done
> timed out waiting for match: (?^:User was holding shared buffer pin for too long) at t/031_recovery_conflict.pl line 318.

No XXX lines this time either, but I've seen then im logfiles that
went through successfully.

> Perhaps this can simply be attributed to the machine being too busy.

With the patches, the problem of dying so often that builds targeting
several distributions in parallel will usually fail is gone.

Christoph