Re: BUG #17485: Records missing from Primary Key index when doing REINDEX INDEX CONCURRENTLY

Петър Славов <pet.slavov@gmail.com>

From: Петър Славов <pet.slavov@gmail.com>
To: Andrey Borodin <x4mmm@yandex-team.ru>
Cc: PostgreSQL mailing lists <pgsql-bugs@lists.postgresql.org>
Date: 2022-05-19T06:23:58Z
Lists: pgsql-bugs
Hi Andrey,
This test looks similar to me but I cannot say what is the difference. My
tests are done under heavier  load maybe.
Also I couldn't reproduce this with a table with low num number of columns
(integrer and text). I am not sure if this is relevant...

Peter


На чт, 19.05.2022 г. в 7:22 ч. Andrey Borodin <x4mmm@yandex-team.ru> написа:

>
>
> > 18 мая 2022 г., в 15:42, PG Bug reporting form <noreply@postgresql.org>
> написал(а):
> > I was able to reproduce this on PostgreSQL 14.1/2/3 locally on
> > docker instance and on AWS EC2.
>
>
> Uhm, that's very...interesting. I'll look closely next week. Though I
> didn't have a chance to reproduce yet.
> We have fixed similar problem in 14.1. And now we have very similar TAP
> test to you reproduction [0]. How do you think, what's the key difference
> between TAP test and your repro?
>
> Thanks! Best regards, Andrey Borodin.
>
> [0]
> https://github.com/postgres/postgres/blob/REL_14_STABLE/contrib/amcheck/t/002_cic.pl

Commits

  1. Revert changes to CONCURRENTLY that "sped up" Xmin advance

  2. Fix possible HOT corruption when RECENTLY_DEAD changes to DEAD while pruning.