On Wed, Dec 13, 2006 at 01:49:08PM -0800, Angva wrote:
> Hi everyone,
>
> First, this group has been good to me, and I thank you guys for the
> valuable help I've found here. I come seeking help with another
> problem. I am not even sure my problem lies in Postgres, but perhaps
> someone here has had a similar problem and could point me in the right
> direction.
As already suggested, we need to see actual error messages.
> I don't know if it means anything, but swap never seems to be used by
> the postgres processes (I stared at top and vmstat for a while), though
> the VIRT column in top definitely shows something. Here are sample
> lines from top while two processed are running:
The figures from top/ps should be taken with a grain of salt, they
include the shared memory which means a lot is being double counted.
> If I run the clusters sequentially, there is no error - they just take
> too long. I was out of ideas so I bounced the server entirely and ran
> the clusters in parallel a few times - success! But I don't want to
> have to bounce the server regularly.
Someone also pointed out overcommit settings. With overcommit turned
off you will get out of memory errors long before it actually runs out.
Perhaps that is the issue? However, we need to see the actual messages
to be sure.
Have a nice day,
--
Martijn van Oosterhout <kleptog@svana.org> http://svana.org/kleptog/
> From each according to his ability. To each according to his ability to litigate.