Re: BUG #17840: Failing to execute auto_explain for logging leads to transaction rollback.

From: "David G(dot) Johnston" <david(dot)g(dot)johnston(at)gmail(dot)com>
To: bokanist(at)gmail(dot)com
Cc: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, pgsql-bugs(at)lists(dot)postgresql(dot)org
Subject: Re: BUG #17840: Failing to execute auto_explain for logging leads to transaction rollback.
Date: 2023-03-15 19:33:46
Message-ID: CAKFQuwa2efAF3VWVVoSQBMUTYx3FsERx2y+o93yToEHL88kwzw@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

On Wed, Mar 15, 2023 at 11:24 AM Sébastien <bokanist(at)gmail(dot)com> wrote:

> What more can I say ? It's clear that the auto_explain feature, which is a
> logging process, can make perfectly fine transactions being rolled back.
> That makes no sense, the transaction should be committed before any logging
> event is triggered, for me it's a bug in the core of postgres. It has
> nothing to do with the plugin used.
>
>
It may be an undesired behavior, and I do sympathize with the position, but
the current design is quite intentional and long-standing and is not going
to be treated as a bug to be fixed immediately and back-patched.

A query you can execute should be incapable of failing if EXPLAIN for the
same query is issued. If some interplay with oracle_fdw causes this
fundamental assumption to be falsified then auto_explain, which works under
that assumption, and oracle_fdw are incompatible and you will need to
decide which one (possibly on a per-transaction basis) you wish to use.

David J.

In response to

Responses

Browse pgsql-bugs by date

  From Date Subject
Next Message Tom Lane 2023-03-15 22:06:25 Re: BUG #17840: Failing to execute auto_explain for logging leads to transaction rollback.
Previous Message Sergei Kornilov 2023-03-15 19:00:18 Re:Certificate expired