Re: Timing issue in pg_basebackup

From: Zsolt Parragi <zsolt(dot)parragi(at)percona(dot)com>
To: Jesper St John <jesper(at)stjohn(dot)se>
Cc: pgsql-bugs(at)lists(dot)postgresql(dot)org, Jesper St John <admin(at)zonezero(dot)se>, Andrey Borodin <x4mmm(at)yandex-team(dot)ru>, Magnus Hagander <magnus(at)hagander(dot)net>, Kirill Reshke <reshkekirill(at)gmail(dot)com>
Subject: Re: Timing issue in pg_basebackup
Date: 2026-11-20 22:59:32
Message-ID: CAN4CZFMcG9v5os=m+NWhgorDt-cCiBBWFwq3atEGoUH24d-0PQ@mail.gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

This seems to break rebuilding cascading standbys: from the
intermediate standby, reusing its existing slot (casc stopped, then
pg_basebackup -h stby -S casc -X stream -R --checkpoint=fast):

master client: stby redo=0/03000028 slot=0/08000060 -> rc=0
patched client: stby redo=0/08000060 slot=0/0D000000 -> rc=1
pg_basebackup: error: backup start point 0/08000060 is older than the
start of the WAL stream at 0/0D000000

And from my previous email:

> and can be invalidated based on max_slot_wal_keep_size

still holds with --checkpoint=false (it works with spread)

And another new issue is that now a failed backup can leave the
created slot on the server:

pg_basebackup -D x -X stream -C -S s1 --manifest-checksums=bogus
pg_basebackup -D x -X stream -C -S s1

--> ERROR: replication slot "s1" already exists

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message surya poondla 2026-11-20 23:47:16 Re: Two issues with REFRESH MATERIALIZED VIEW CONCURRENTLY
Previous Message Tom Lane 2026-11-20 22:47:40 Re: BUG #19924: unrecognized node type in create_plan_recurse: UNION, LIMIT, json_populate_recordset, inheritance