Waldomiro wrote:
Is It possible the checkpoint is doing that? Or the archiving? How can
I see?
If you're using PostgreSQL 8.3 or later, you can turn on log_checkpoints
and you'll get a note when each checkpoint finishes. The parts that are
more likely to slow the server down are right at the end, so if you see
a bunch of slow queries around the same time as the checkpoint message
appears in the logs, that's the likely cause. Bad checkpoint behavior
can certainly cause several seconds of freezing on a system with 32GB of
RAM, because with that much data you can have quite a bit in the OS
write cache that all gets forced out at the end of the checkpoint.
Finding when the checkpoints happen on 8.2 or earlier is much harder; I
can tell you what to look for on Linux for example, but it's kind of
painful to track them down.
--
Greg Smith 2ndQuadrant Baltimore, MD
PostgreSQL Training, Services and Support
greg@xxxxxxxxxxxxxxx www.2ndQuadrant.com
--
Sent via pgsql-performance mailing list (pgsql-performance@xxxxxxxxxxxxxx)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-performance
