Horton

Re: ERROR: subtransaction logged without previous top-level txn record

Dan Katz <dkatz@joor.com>

View thread View on postgr.es
From: Dan Katz <dkatz@joor.com>
To: Arseny Sher <a.sher@postgrespro.ru>
Cc: Andres Freund <andres@anarazel.de>, Alvaro Herrera <alvherre@2ndquadrant.com>, "Hsu, John" <hsuchen@amazon.com>, "pgsql-bugs@lists.postgresql.org" <pgsql-bugs@lists.postgresql.org>
Date: 2020-01-30T20:09:57Z
Lists: pgsql-bugs, pgsql-hackers 
Arseny,

I was hoping you could give me some insights about how this bug might
appear with multiple replications slots. For example if I have two
replication slots would you expect both slots to see the same error, even
if they were started, consumed or the LSN was confirmed-flushed at
different times?

Dan

On Tue, Dec 17, 2019 at 10:15 AM Arseny Sher <a.sher@postgrespro.ru> wrote:

>
> Dan Katz <dkatz@joor.com> writes:
>
> > Is the resolution of the issue in this thread being tracked elsewhere,
> > either in a commit fest or other stream of work?
>
> Ok, I've created a cf entry:
> https://commitfest.postgresql.org/26/2383/
>
> I believe it is the most important to commit at least
>
> 0002-Stop-demanding-that-top-xact-must-be-seen-before-sub.patch
>
> from my mail above -- as we see, this issue creates problems in the
> field. Moreover, the patch is trivial and hopefully I've shown that
> ERROR triggers were spurious and there is no easy way to relax the
> check.
>
> --
> Arseny Sher
> Postgres Professional: http://www.postgrespro.com
> The Russian Postgres Company
>

Commits

  1. Stop demanding that top xact must be seen before subxact in decoding.

  2. logical decoding: process ASSIGNMENT during snapshot build

  3. Fix operator naming in pg_trgm GUC option descriptions