On Mon, 04 Jul 2011 09:56:55 +0200 Erwan <[email protected]> wrote: > Hi there, > > I have the following use case: > Sometimes I have little timeouts on my network and pgpool-II detaches > the node which is no more reachable (the state is equal to 3). > I wonder if it was possible to reattach "auto-magically" this node when > there is no more timeout?
Pgpool does not reattach automatically except in raw mode at present. > I imagine the health check is a solution to do that but I have several > tiny dumb questions about this check: > - how the check is done? Because we have to specify a user but no > password... I imagine I also have to change the pg_hba.conf file You don't have to change the pg_hba.conf. Pgpool's healthcheck sends StartupMessage which is based on PostgreSQL protocol to the backends, and it checks the response from the backends. Pgpool judges it's active if receive some response. So you do NOT need to change the authentication method to "trust". > - does it allow pgpool-II to reattach a detached backend? No, the helthcheck never reattach automatically. -- Toshihiro Kitagawa SRA OSS, Inc. Japan _______________________________________________ Pgpool-general mailing list [email protected] http://pgfoundry.org/mailman/listinfo/pgpool-general
