| From: | Fujii Masao <masao(dot)fujii(at)gmail(dot)com> |
|---|---|
| To: | Manu <manuelreyesbravo(at)gmail(dot)com> |
| Cc: | Ayush Tiwari <ayushtiwari(dot)slg01(at)gmail(dot)com>, Chao Li <li(dot)evan(dot)chao(at)gmail(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Incremental backup can use a summary past a timeline switch |
| Date: | 2026-10-30 10:12:36 |
| Message-ID: | CAHGQGwHgE=cC6K-jQzD4EtE5n_GPSGFOt3G3JAEiZggAyA1jTA@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Sat, Oct 24, 2026 at 4:43 AM Manu <manuelreyesbravo(at)gmail(dot)com> wrote:
> Without v1-0001 it was 5 of
> 5 on btrfs and 0 of 5 on tmpfs, where readdir happens to merge the
> old-timeline summary first.
Thanks for testing! The patch looks good to me.
I'm thinking to commit this before the fix for [1], since changing
the merge order could expose the bug in cases where it was previously
hidden. Barring any objections, I'll commit this patch first.
For v20, we could also consider sharing the filtering logic between
GetWalSummaries() and FilterWalSummaries() to avoid similar inconsistencies.
Regards,
[1] https://postgr.es/m/19774-8a1cbfd1774fcd0c@postgresql.org
--
Fujii Masao
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Nitin Jadhav | 2026-10-30 10:13:44 | Re: Reporting WAL replay progress during pre-consistency standby reovery |
| Previous Message | vignesh C | 2026-10-30 09:55:45 | Re: Avoid resetting subscription two_phase state to pending when already enabled |