Horton

Re: [PATCH] Clarify the behavior of the system when approaching XID wraparound

Robert Haas <robertmhaas@gmail.com>

View thread View on postgr.es
From: Robert Haas <robertmhaas@gmail.com>
To: Aleksander Alekseev <aleksander@timescale.com>
Cc: PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>, Peter Geoghegan <pg@bowt.ie>, Pavel Borisov <pashkin.elfe@gmail.com>, Peter Eisentraut <peter@eisentraut.org>
Date: 2023-10-16T18:06:01Z
Lists: pgsql-hackers

Attachments

On Fri, Oct 13, 2023 at 5:03 AM Aleksander Alekseev
<aleksander@timescale.com> wrote:
> > Thanks for working on this. I will be relieved once this is finally
> > taken care of.
>
> +1, and many thanks for your attention to the patchset and all the details!

Cool. I committed that and back-patched to v14, and here's the rest.
0001 makes the terminology change that I proposed earlier, and then
0002 is the remainder of what was in the previous patch set that
wasn't covered by what I committed already, with a few adjustments.

In particular, I preferred to stick with "avoid" rather than changing
to "prevent," and I thought it was clearer to refer to "failures"
plural rather than "failure" collective. These are arguable decisions,
though.

I propose to commit these changes only to master. I have included a
fairly long paragraph about that in the commit message for 0002.

-- 
Robert Haas
EDB: http://www.enterprisedb.com

Commits

  1. Reword messages about impending (M)XID exhaustion.

  2. Talk about assigning, rather than generating, new MultiXactIds.

  3. Update the documentation on recovering from (M)XID exhaustion.