Creating backup history files for backups taken from standbys

From: Michael Paquier <michael(dot)paquier(at)gmail(dot)com>
To: PostgreSQL mailing lists <pgsql-hackers(at)postgresql(dot)org>
Cc: Robert Haas <robertmhaas(at)gmail(dot)com>
Subject: Creating backup history files for backups taken from standbys
Date: 2017-08-10 06:56:30
Message-ID: CAB7nPqQvVpMsqJExSVXHUwpXFRwojsb-jb4BYnxEQbjJLfw-yQ@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox
Thread:
Lists: pgsql-hackers

Hi all,

In the recent thread related to a bug in pg_stop_backup when waiting
for WAL segments to be archived, it has been mentioned that it would
be nice to get backup history files generated as well on standbys:
https://www.postgresql.org/message-id/CA+Tgmobh3=quCzugFP5yA3-z5uzz=YKZ7arnoApJoPN7=0H8TQ@mail.gmail.com

Backup history files are not essential to backups, but they are for
debugging purposes, and it has bugged me hard that we should have a
consistent behavior in this area, particularly because with
archive_mode = always they can be archived by a standby.

The previous thread has ended with 52f8a59d, which addressed the
original problem but discarded the backup history file generation
during recovery. Attached is a patch to fix this inconsistency, parked
in next CF. I am not arguing for this change as a bug fix, but as an
improvement for PG11.

Thoughts?
--
Michael

Attachment Content-Type Size
standby-backup-history.patch application/octet-stream 4.1 KB

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Michael Paquier 2017-08-10 07:22:23 Re: More race conditions in logical replication
Previous Message Andreas Seltenreich 2017-08-10 06:50:59 Re: Server crash (FailedAssertion) due to catcache refcount mis-handling