Horton

Re: Use of backup_label not noted in log

David Steele <david@pgmasters.net>

View thread View on postgr.es
From: David Steele <david@pgmasters.net>
To: Michael Paquier <michael@paquier.xyz>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Michael Banck <mbanck@gmx.net>, Andres Freund <andres@anarazel.de>, Pg Hackers <pgsql-hackers@postgresql.org>, Laurenz Albe <laurenz.albe@cybertec.at>
Date: 2024-01-29T14:03:19Z
Lists: pgsql-hackers 
On 1/28/24 20:09, Michael Paquier wrote:
> On Fri, Jan 26, 2024 at 12:08:46PM +0900, Michael Paquier wrote:
>> Well, I'm OK with this consensus on 1d35f705e if folks think this is
>> useful enough for all the stable branches.
> 
> I have done that down to REL_15_STABLE for now as this is able to
> apply cleanly there.  Older branches have a lack of information here,
> actually, because read_backup_label() does not return the TLI
> retrieved from the start WAL segment, so we don't have the whole
> package of information.

I took a pass at this on PG14 and things definitely look a lot different 
back there. Not only is the timeline missing, but there are two sections 
of code for ending a backup, one for standby backup and one for primary.

I'm satisfied with the back patches as they stand, unless anyone else 
wants to have a look.

Regards,
-David

Commits

  1. Add more LOG messages when starting and ending recovery from a backup