I'm sure I'm missing something obvious here, but we're having a bit of a
problem with pgpool-II. We have pgpool configured with num_init_children
= 100 and max_pool = 1. We have two backends each with max_connections =
150. Today we started receiving errors from postgres about exceeding the
max number of connections. We noticed that one of our backends was
apparently giving pgpool trouble, so we detached it. The error we were
getting was:
2010-08-13 08:09:04 ERROR: pid 759: pool_read_kind: kind does not match
between master(45) slot[1] (53)
2010-08-13 08:09:04 ERROR: pid 759: pool_do_auth: failed to read kind
before BackendKeyData
detaching the node didn't solve our connection problems, so we detached
the last node, shut it down, increased max connections to 1000 and
brought it back up. Now everything is working, but we've noticed that
pgpool has 300 processes waiting for, or handling connections. I assume
this is why we wound up exceeding our max connections, my question is
why do we have so many pgpool children when the max is set to 100?
-Tevis Money
_______________________________________________
Pgpool-general mailing list
[email protected]
http://pgfoundry.org/mailman/listinfo/pgpool-general
