Обсуждение: Full backup failed

On Mon, 2024-02-19 at 09:35 +0530, Rajesh Kumar wrote:
> 6time="2024-02-18T18:26:59Z" level=info msg="stderr=[ERROR: [082]: WAL segment 0000003F0000050A0000000C was not
archivedbefore the 60000ms timeout\n HINT: check the archive_command to ensure that all options are correct (especially
--stanza).\nHINT: check the PostgreSQL server log for errors.\n HINT: run the 'start' command if the stanza was
previouslystopped.\n]" 

Look into the PostgreSQL log for messages from the "archive_comand".

Yours,
Laurenz Albe

Hi,

On Mon, Feb 19, 2024 at 01:13:36PM +0530, Rajesh Kumar wrote:
> Archive command is set to pgbackrest --stanza=db archive-push "%p"

That looks right.
 
> There were no such issue since last year.
> 
> In postgres logs, don't see anything related to archive. But I see,
> WARNING,0100, aborting backup due to backend exiting before pg_backup_stop
> was called....pgbackrest[backup] client backend, 58568621736954596

This error is due to the timeout that pgbackrest backend suffered.

I think one reason pgbackrest backup could time out after 60s is WAL
pileup, i.e. it tries to check whether the current WAL archive was
successfully archived in 60s but if there is a huge lag in archiving
(for whatever reason) the current WAL segment might not get archived for
another 60 seconds and the backup fails.

So if you have that data in monitoring, check whether archiving was
maybe lagging behind, or otherwise, check whether there are a lot of
.ready files in the pg_wal/archive_status sub-directory right now and/or
possibly just try the backup again.


Michael