Date: Thu, 14 Jul 2005 15:08:30 +1000
To: pgsql-performance@postgresql.org
Subject: lots of updates on small table
Message-ID: <42D5F34E.nail9QE1195IE@pluto.mirrabooka.com>
User-Agent: nail 11.6 9/7/04
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: alison@mirrabooka.com (Alison Winters)

Hi,

Our application requires a number of processes to select and update rows
from a very small (<10 rows) Postgres table on a regular and frequent
basis.  These processes often run for weeks at a time, but over the
space of a few days we find that updates start getting painfully slow.
We are running a full vacuum/analyze and reindex on the table every day,
but the updates keep getting slower and slower until the processes are
restarted.  Restarting the processes isn't really a viable option in our
24/7 production environment, so we're trying to figure out what's
causing the slow updates.

The environment is as follows:

Red Hat 9, kernel 2.4.20-8
PostgreSQL 7.3.2
ecpg 2.10.0

The processes are all compiled C programs accessing the database using
ECPG.

Does anyone have any thoughts on what might be happening here?

Thanks
Alison