[pgpool-general: 8455] Re: pgpool_status and backend status update rules

[Tatsuo Ishii]

Hi,

> Hello,
> from what I understood, the /var/log/postgresql/pgpool_status file contains
> a text line for each of the backend servers. The line text is "up" or
> "down" based on the availability of the server.

Basically correct. Actually there is one more status "unused" which is
set when a backend is not properly configured.

> It seems to be that pgpool load the file while starting up and set the
> status of the backends according to the file. Then, when the connection to
> a backend fails, I suppose it will update the file. Or, maybe, when
> pcp_detach_node is run.

Again, correct.

> May pgpool also sets the status "down" for a
> standby backend that has a non zero delay_threshold?

No.

> I am unsure about when pgpool sets again "up" a backend status. Is that
> only when pcp_attach_node is called?

Other than pcp_attach_node, auto_failback
(https://www.pgpool.net/docs/latest/en/html/runtime-config-failover.html#RUNTIME-CONFIG-FAILOVER-SETTINGS)
automatically attaches a standby node once it becomes healthy again.

Best reagards,
--
Tatsuo Ishii
SRA OSS LLC
English: http://www.sraoss.co.jp/index_en/
Japanese:http://www.sraoss.co.jp