Re: Re: Assertion failure in WAL summarizer

From: Raja Sai pranav <rajasaipranav0(at)gmail(dot)com>
To: zhong950419(at)gmail(dot)com
Cc: robertmhaas(at)gmail(dot)com, manuelreyesbravo(at)gmail(dot)com, masaofujii(at)gmail(dot)com, darianlepikhova(at)gmail(dot)com, ayushtiwarislg01(at)gmail(dot)com, pgsql-hackers(at)lists(dot)postgresql(dot)org
Subject: Re: Re: Assertion failure in WAL summarizer
Date: 2026-11-11 04:00:05
Message-ID: CAPHduZANUbs3cUCstL=2GMqWJPoZf3xn4NoXyF7is8bHeobccg@mail.gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi,

Beyond confirming v4 applies and the suites pass, I checked that each new
test
actually reproduces its failure on unpatched master -- i.e. with the three
test
files in place but the code changes reverted:

- pg_walsummary/t/
https://www.google.com/url?q=http://004_promote_past_target.pl&source=gmail&ust=1794456004508000&sa=E:
fails at "the summarizer does not
read past replay", and pg_ctl promote then exits 1 -- the post-promotion
path
this series fixes.
- pg_combinebackup/t/
https://www.google.com/url?q=http://015_summary_past_switch.pl&source=gmail&ust=1794456004508000&sa=E:
the same "reads past replay"
failure on the standby, and promote exits 1.
- pg_combinebackup/t/
https://www.google.com/url?q=http://016_truncation_after_target.pl&source=gmail&ust=1794456004508000&sa=E:
the server terminates
abnormally during the post-promotion incremental backup
(pg_basebackup: server closed the connection unexpectedly).

With all three patches applied, every test passes (pg_combinebackup 147/147,
pg_walsummary 34/34, on an assert-enabled build). So the tests reproduce the
reported failures -- including the server crash on the
truncation-after-target
path -- and the series resolves them.

Regards,
Sai Pranav

On Wed, Nov 11, 2026 08:04 AM, shihao zhong <zhong950419(at)gmail(dot)com> wrote:

> Hi,
>
> v3 no longer applies on master. I did not see any new comments.
>
> So v4 is just a rebase.
>
> Thanks,
>
>
>
>
>

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message João Rito 2026-11-11 04:11:13 Re: Small fixes needed by high-availability tools
Previous Message shihao zhong 2026-11-11 03:25:00 Re: Set 1s WaitLatch timeout if standby limit has expired in ResolveRecoveryConflictWithBufferPin