[pgpool-general: 7524] Re: Recovery after failure on a single backend

[Emond Papegaaij]

Hi Tatsuo,

Yes, that's also what I got from the documetation. Thanks for the
clarification. For our single-node setup we went for
DISALLOW_TO_FAILOVER. For our cluster setup, we continue to use
auto_failback.

Best regards,
Emond

On Sat, Apr 24, 2021 at 12:23 AM Tatsuo Ishii <ishii at sraoss.co.jp> wrote:
>
> >>> 2) Enable auto_failback. Unlike #1, failover completes. But after some
> >>> period, the backend automatically becomes online again if connection
> >>> to the backend becomes healthy again.
> >>
> >> The documentation states that this option relies on streaming
> >> replication. We use this option in our cluster setup to automatically
> >> reattach backends when streaming is in sync again. I don't think it
> >> will work with a single backend because pgpool has no way of detecting
> >> that the connection is healthy again. The health checks don't seem to
> >> run on detached nodes.
> >
> > You are right. Sorry for wrong suggestion.
>
> One correction. The reason why auto_failback does not work on a single
> node is, because the node is not a streaming replication standby.
> (Note that auto_failback runs on detached node.)
>
> Best regards,
> --
> Tatsuo Ishii
> SRA OSS, Inc. Japan
> English: http://www.sraoss.co.jp/index_en.php
> Japanese:http://www.sraoss.co.jp