| From: | Benoit Lobréau <benoit(dot)lobreau(at)dalibo(dot)com> |
|---|---|
| To: | David Steele <david(at)pgbackrest(dot)org>, "Andrey M(dot) Borodin" <x4mmm(at)yandex-team(dot)ru> |
| Cc: | Pg Hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Fix logging for invalid recovery timeline |
| Date: | 2025-02-19 09:51:49 |
| Message-ID: | 4ed10bc8-5513-4d8e-8643-8abcaa08336d@dalibo.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Hi,
I think the changes make sense. Would it be helpful to add the origin of
the checkpoint record we are referring to ? (i.e. control file or backup
label).
For example:
* Latest checkpoint in the control file is at %X/%X on timeline %u,
* Checkpoint location in the backup_label file is at %X/%X on timeline %u,
The current message refers to "Latest checkpoint" which automatically
makes me think about the control file (specifically, the pg_controldata
output).
Now that I have read the code, I understand which one we are referring
to, but for someone who hasn't (or me in a few month) it might be useful ?
--
Benoit Lobréau
Consultant
http://dalibo.com
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Bertrand Drouvot | 2025-02-19 09:56:18 | Re: POC: enable logical decoding when wal_level = 'replica' without a server restart |
| Previous Message | Alexander Korotkov | 2025-02-19 09:48:04 | Re: Improve statistics estimation considering GROUP-BY as a 'uniqueiser' |