From: | "Peter Darley" <pdarley(at)Kinesis-CEM(dot)com> |
---|---|
To: | "Pgsql-Admin" <pgsql-admin(at)postgresql(dot)org> |
Subject: | Postgresql optimization |
Date: | 2001-11-13 15:35:57 |
Message-ID: | NNEAICKPNOGDBHNCEDCPMEBLCBAA.pdarley@kinesis-cem.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
Friends,
I have what seems to me to be a fairly beefy Linux box (2x1Ghz cpu, 768meg
memory) that is a dedicated Postgres server, and which hasn't been
performing the way I felt it should.
I found a good article on the web about optimizing Postgres on Linux, (
http://www.phpbuilder.com/columns/smith20010821.php3?print_mode=1 )and the
thing that seems to be the bottleneck for my database appears to be the
amount of memory that Linux sets aside for shared memory space. I set my
machine up like the guy in the article (by doing echo 128000000 >
/proc/sys/kernel/shmmax to set the shared memory to 128 megs), and things
are going way faster now than they were. I'm still at 340 megs free on the
machine, and I would like to bump this up by half again (192 megs), and then
up to double (256 megs) if that works out, however I don't really understand
all the ramification for this, so I just wanted to check with y'all and see
if you thought I would be getting myself in trouble...
Thanks,
Peter Darley
From | Date | Subject | |
---|---|---|---|
Next Message | Nick Fankhauser | 2001-11-13 19:55:53 | PostgreSQL->JDBC->Tomcat->Apache resource uses |
Previous Message | Tom Lane | 2001-11-12 23:43:27 | Re: Very long COMMIT times? |