Re: Deathly slow performance on SMP red-hat system

["Steve Poe" <steve(dot)poe(at)gmail(dot)com>]

Patrick,

Have you modified your postgresql.conf file or is this default out-of-the-box too? Are the database and database logs (pg_xlog) on shared disc space?

Steve

On 8/28/06, Patrick TJ McPhee <ptjm(at)interlog(dot)com> wrote:

I have a Red Hat ELsmp system running on a dual opteron hp server with
16 gigs of memory. I believe the RH installation is straight out of the
box. I've compiled postgres 8.1.4 from sources.

The problem is that query performance is horrible, and I can't think
why, although it seems clear the problem is outside the query engine.
For instance, "select 1" takes on the order of 100ms. Explain analyze
doesn't shed much light on why this might be:

explain analyze select 1;
                                     QUERY PLAN
------------------------------------------------------------------------------------
Result  (cost=0.00..0.01 rows=1 width=0) (actual time=0.005..0.006 rows=1 loops=1)
Total runtime: 0.059 ms
(2 rows)

But if I turn on duration logging, I get timings like
LOG:  duration: 91.480 ms

I was able to pg_reload 80 gigs of data in a reasonable amount of time, but
individual queries are taking a long time.

I'm wondering if anyone's seen anything like this.
--

Patrick TJ McPhee
North York  Canada
ptjm(at)interlog(dot)com

---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
       choose an index scan if your joining column's datatypes do not
       match