Pg9.2: Monitoring NOTIFY queue

From: Jiří Hlinka <jiri(dot)hlinka(at)gmail(dot)com>
To: "pgsql-admin(at)postgresql(dot)org" <pgsql-admin(at)postgresql(dot)org>
Cc: Jiří Hlinka <jiri(dot)hlinka(at)gmail(dot)com>
Subject: Pg9.2: Monitoring NOTIFY queue
Date: 2014-12-01 14:54:14
Message-ID: CADCCO5q8UUEeDQCFQkbm_mGwVFZZfnG7nJ1EF1DH4AiXuMvY5Q@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-admin

Hello,

is there any way how to monitor NOTIFY's queue? I'm talking about the queue
described here in the "Notes" part:
http://www.postgresql.org/docs/9.2/static/sql-notify.html

Recently there were WARNING messages in DB logs like this:
WARNING: NOTIFY queue is 55% full
DETAIL: The server process with PID <PID number> is among those with the
oldest transactions.
HINT: The NOTIFY queue cannot be emptied until that process ends its
current transaction.

... in coincidence we stopped all processess connected to DB so the problem
was solved as a side effect to this step, unfortunately we are unable to
find which process (described by PID in log messages above) was cause this
filling of NOTIFY's queue.

We need to monitor state of notify queue so we would be able to solve the
problem before the queue fills up. There seems to be a problem in some
application that causes this behavior but right now we are unable to say
which one and therefore to solve the problem with the specific development
team.

Thank you,
Jiri

Browse pgsql-admin by date

  From Date Subject
Next Message Denniss 2014-12-02 10:32:48 Re: Migration from Sybase to Postgresql
Previous Message Doiron, Daniel 2014-11-26 19:34:55 Re: [ADMIN] Active/Active clustering in postgres