[pgpool-general: 1514] Re: reset_query_list DISCARD ALL queries getting stuck

Gavin Harcourt gh at tucasi.com
Wed Mar 20 00:39:11 JST 2013

As always I realise I'm being a bit stupid just as I hit send on an email.
The DISCARD ALL command had completed, the session however was remaining
open blocking any DROP DATABASE command I issued. Of course the solution to
this is to set a timeout on the connection session within pgpool.conf so
that the open session closes and the database can be safely dropped.


-----Original Message-----
From: pgpool-general-bounces at pgpool.net
[mailto:pgpool-general-bounces at pgpool.net] On Behalf Of Gavin Harcourt
Sent: 19 March 2013 15:31
To: pgpool-general at pgpool.net
Subject: [pgpool-general: 1513] reset_query_list DISCARD ALL queries getting


I've been trying to find any mention of this particular problem but it
doesn't seem to have popped up before so I was hoping someone on this list
might be able to help shed some light on it.

I'm running pg-pool 3.2.3 with Postgresql 9.2.3 in load balancing and
connection pooling modes to 2 servers (using the postgres hot standby
replication between them). I've got the requests being load balanced
correctly with read only requests going to the standby server and write
requests going to the primary server. Whenever a connection is closed to
pgpool the DISCARD ALL command issued by pg_pool (as defined in
reset_query_list in the pgpool.conf) is getting stuck and never completing.
I can see no errors in the log files  for either Postgresql server and no
resource usage on the servers. If I submit the command myself it completes
immediately as you'd expect.

This is causing a problem as the sessions to the database never complete
meaning that I can never reuse that connection or perform a DROP DATABASE
command. Does anyone have any suggestions on where to even start looking
with this?

Gavin Harcourt

pgpool-general mailing list
pgpool-general at pgpool.net

More information about the pgpool-general mailing list