VS: BUG #17218: Cluster recovery is not working properly.

Hi,

Thanks for the information.

Regards,

Reijo Suhonen

Lähettäjä: Magnus Hagander <magnus(at)hagander(dot)net>
Lähetetty: maanantai 11. lokakuuta 2021 13.12
Vastaanottaja: Suhonen Reijo (Valtori) <reijo(dot)suhonen(at)valtori(dot)fi>; PostgreSQL mailing lists <pgsql-bugs(at)lists(dot)postgresql(dot)org>
Aihe: Re: BUG #17218: Cluster recovery is not working properly.

On Mon, Oct 11, 2021 at 12:09 PM PG Bug reporting form <noreply(at)postgresql(dot)org<mailto:noreply(at)postgresql(dot)org>> wrote:
The following bug has been logged on the website:

Bug reference: 17218
Logged by: Reijo Suhonen
Email address: reijo(dot)suhonen(at)valtori(dot)fi<mailto:reijo(dot)suhonen(at)valtori(dot)fi>
PostgreSQL version: 14.0
Operating system: Red Hat Enterprise Linux Server 7.9 (Maipo)
Description:

Restoring a cluster to a specific point in time does not work. The restore
ends with the error message "FATAL, XX000," recovery ended before configured
recovery target was reached " and the cluster goes down.
The same test with version 12 works normally.

This doesn't look like a bug. You have specified a recovery target (09:28) that's beyond the end of what's in WAL (09:27:20). You shouldn't do that -- if you want to recover until the end of WAL, you shouldn't specify a target at all.

The behaviour of this case changed in PostgreSQL 13 which is probably why you didn't see it before. https://www.postgresql.org/docs/release/13.0/ has the entry "Generate an error if recovery does not reach the specified recovery target (Leif Gunnar Erlandsen, Peter Eisentraut) Previously, a standby would promote itself upon reaching the end of WAL, even if the target was not reached.".

--
Magnus Hagander
Me: https://www.hagander.net/<http://www.hagander.net/>
Work: https://www.redpill-linpro.com/<http://www.redpill-linpro.com/>