Hello Michael,
03.03.2026 08:13, Michael Paquier wrote:
+ autovacuum = off
+ checkpoint_timeout = 1h
Why do we need these? An explanation seems in order in the shape of a
commit, or these should be removed.
Is there a different trick than the one posted at [1] to check the
stability of the proposal? I am wondering if I am missing something,
or if that's all. Alexander?
[1]: https://www.postgresql.org/message-id/e1cf52d2-c344-4dfd-a918-e5f20ff04fa2@gmail.com
There was also a separate test (independent of bgwriter, autovacuum,
checkpointer,...) at the top of the thread: [1].
FWIW, 004_timeline_switch also failed the past month in the same way [2].
I've just tested 012_subtransactions.pl with the script I proposed at [3]
and it passed 100 iterations (because it performs $node_primary->stop;
before $node_standby->promote;, I suppose).
[1]
https://www.postgresql.org/message-id/b0102688-6d6c-c86a-db79-e0e91d245b1a%40gmail.com [2]
https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=mandrill&dt=2026-02-07%2006%3A33%3A20 [3]
https://www.postgresql.org/message-id/045cab6f-4738-417e-b551-01adba44d6c3%40gmail.com Best regards,
Alexander
