Horton

Re: BF mamba failure

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Bertrand Drouvot <bertranddrouvot.pg@gmail.com>
Cc: Kouber Saparev <kouber@gmail.com>, pgsql-hackers@lists.postgresql.org
Date: 2024-11-15T04:45:24Z
Lists: pgsql-hackers 
On Thu, Nov 14, 2024 at 10:07:41AM +0000, Bertrand Drouvot wrote:
> === 1
> 
> Maybe use "generation" instead of generation in the comments (where it's not done,
> some comments do it already).

I've tweaked things to be more consistency, and applied that down to 15.

> === 2
> 
> We could think about adding a test. That should be doable with replication slots
> or custom pgstats kinds and probably injection points. But I'm not sure that's
> worth it, as the code in the patch looks "simple" enough. Thoughts?

The tricky part is that this path involves a backend shutdown.  We
should be able to do something with a wait before the dshash_find() in
pgstat_release_entry_ref() with a BackgroundPsql object that gets
stopped, but I'm not sure if we can guarantee a lookup of
pg_stat_activity at this stage.  Let me see..
--
Michael

Commits

  1. Add information about "generation" when dropping twice pgstats entry

  2. Fix race conditions with drop of reused pgstats entries