From: | Greg Smith <gsmith(at)gregsmith(dot)com> |
---|---|
To: | John Huttley <John(at)mib-infotech(dot)co(dot)nz> |
Cc: | pgsql-performance(at)postgresql(dot)org |
Subject: | Re: Slow updates, poor IO |
Date: | 2008-09-27 07:04:08 |
Message-ID: | Pine.GSO.4.64.0809270301290.27800@westnet.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-performance |
On Sat, 27 Sep 2008, John Huttley wrote:
> I've got 32M shared on a 1G machine and 16 checkpoint segments.
> I'll run some tests against 64 segments and see what happens.
Increase shared_buffers to 256MB as well. That combination should give
you much better performance with the type of update you're doing. Right
now the database server has to write the index blocks updated to disk all
the time because it has so little working room to store them in. If an
index block is updated but there is room to keep it memory, it doesn't
have to get written out, which considerably lowers the overhead here.
--
* Greg Smith gsmith(at)gregsmith(dot)com http://www.gregsmith.com Baltimore, MD
From | Date | Subject | |
---|---|---|---|
Next Message | Scott Marlowe | 2008-09-27 15:09:09 | Re: Slow updates, poor IO |
Previous Message | John Huttley | 2008-09-26 23:09:31 | Re: Slow updates, poor IO |