[pgpool-general: 3511] Problem with show pool_nodes

[Christophe Le Roux]

Hello,

I've set up a postgresql and pgpool cluster on 2 nodes.
Postgresql instances are 9.3.6 and use streaming replication.
Pgpool are in HA mode.

You can see my pgpool config and pgpool.log in attachment.

When I shutdown the primary node, or if I promote the standby with pcp_promote_node at first show pool_nodes give good values, but after some seconds, it display  2 standby nodes.
If I query pg_is_in_recovery()with the user configured in sr_check_user one is master and the other is standby (the failover process do his job).
After a pgpool restart I have good value.

Can you help me to solve this problem ?

Cordialement,

Christophe Le Roux
Database engineer
christophe.le.roux at fr.clara.net<mailto:christophe.le.roux at fr.clara.net>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.pgpool.net/pipermail/pgpool-general/attachments/20150312/2e37caf6/attachment.htm>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: pgpool-config.txt
URL: <http://www.pgpool.net/pipermail/pgpool-general/attachments/20150312/2e37caf6/attachment.txt>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pgpool.log
Type: application/octet-stream
Size: 198856 bytes
Desc: pgpool.log
URL: <http://www.pgpool.net/pipermail/pgpool-general/attachments/20150312/2e37caf6/attachment.obj>