Interesting - in this quick snapshot there is no I/O happening at all.
What happens when you track the activity for a longer period of time?
How about just capturing vmstat during a period when the queries are
slow?
Has the load average been this high forever or are you experiencing a
growth in workload? 463 processes all doing CPU work will take 100x as
long as one query on a 4 CPU box, have you worked through how long you
should expect the queries to take?
- Luke
> -----Original Message-----
> From: Willo van der Merwe [mailto:willo@studentvillage.co.za]
> Sent: Wednesday, August 30, 2006 4:35 AM
> To: Luke Lonergan
> Cc: Merlin Moncure; pgsql-performance@postgresql.org
> Subject: Re: [PERFORM] PostgreSQL performance issues
>
> Luke Lonergan wrote:
> >> Currently the load looks like this:
> >> Cpu0 : 96.8% us, 1.9% sy, 0.0% ni, 0.3% id, 0.0% wa,
> 0.0% hi,
> >> 1.0% si
> >> Cpu1 : 97.8% us, 1.6% sy, 0.0% ni, 0.3% id, 0.0% wa,
> 0.0% hi,
> >> 0.3% si
> >> Cpu2 : 96.8% us, 2.6% sy, 0.0% ni, 0.3% id, 0.0% wa,
> 0.0% hi,
> >> 0.3% si
> >> Cpu3 : 96.2% us, 3.2% sy, 0.0% ni, 0.3% id, 0.0% wa,
> 0.0% hi,
> >> 0.3% si
> >>
> >
> > All four CPUs are hammered busy - check "top" and look for runaway
> > processes.
> >
> > - Luke
> >
> >
> >
> Yes, the first 463 process are all postgres. In the meanwhile
> I've done:
> Dropped max_connections from 500 to 250 and Upped
> shared_buffers = 50000
>
> Without any apparent effect.
>
>