Horton

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

Pavel Borisov <pashkin.elfe@gmail.com>

View thread View on postgr.es
From: Pavel Borisov <pashkin.elfe@gmail.com>
To: John Naylor <john.naylor@enterprisedb.com>
Cc: Aleksander Alekseev <aleksander@timescale.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2023-04-04T09:57:45Z
Lists: pgsql-hackers

Attachments

Hi!

I've looked into the patches v4.
For 0001:
I think long "not accepting commands that generate" is equivalent to
more concise "can't generate".
For 0003:
I think double mentioning of Vacuum at each errhist i.e.: "Execute a
database-wide VACUUM in that database" and "...or run a manual
database-wide VACUUM." are redundant. The advice "Ensure that
autovacuum is progressing,..." is also not needed after advice to
"Execute a database-wide VACUUM in that database".

For all:
In a errhint's list what _might_ be done I think AND is a little bit
better that OR as the word _might_ means that each of the proposals in
the list is a probable, not a sure one.

The proposed changes are in patchset v5.

Kind regards,
Pavel Borisov,
Supabase.

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.