Re: [BUG] recovery of prepared transactions during promotion can fail

From: Michael Paquier <michael(at)paquier(dot)xyz>
To: Kyotaro Horiguchi <horikyota(dot)ntt(at)gmail(dot)com>
Cc: julian(dot)markwort(at)cybertec(dot)at, pgsql-hackers(at)postgresql(dot)org
Subject: Re: [BUG] recovery of prepared transactions during promotion can fail
Date: 2023-06-19 07:27:27
Message-ID: ZJADX9pvRpmX8i4G@paquier.xyz
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Mon, Jun 19, 2023 at 02:41:54PM +0900, Kyotaro Horiguchi wrote:
> At Mon, 19 Jun 2023 14:24:44 +0900, Michael Paquier <michael(at)paquier(dot)xyz> wrote in
>> On Fri, Jun 16, 2023 at 04:27:40PM +0200, Julian Markwort wrote:
>>> I've attached patches for PG 14 and PG 13 that mimic the change in
>>> PG15 (commit 811051c ) and reorder the crucial events, placing the
>>> recovery of prepared transactions *before* renaming the file.
>>
>> Yes, I think that's OK. I would like to add two things to your
>> proposal for all the existing branches.
>> - Addition of a comment where RecoverPreparedTransactions() is called
>> at the end of recovery to tell that we'd better do that before working
>> on the last partial segment of the old timeline.
>> - Enforce the use of archiving in 009_twophase.pl.
>
> Both look good to me.

Okay, cool. Thanks for double-checking, so let's do something down to
13, then..
--
Michael

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Juan José Santamaría Flecha 2023-06-19 07:42:04 Re: Inconsistent results with libc sorting on Windows
Previous Message Ronan Dunklau 2023-06-19 07:01:56 Re: Use generation context to speed up tuplesorts