Re: PostgreSQL 9.2 benchmarks

[Mindaugas Rasiukevicius <rmind@xxxxxxxxxx>]

Hello Francois,

I am CC-ing tech-kern as well.

Francois Tigeot <ftigeot@wolfpond.org> wrote:
> On Mon, Aug 27, 2012 at 11:31:30AM +0100, Mindaugas Rasiukevicius wrote:
> > 
> > Francois Tigeot <ftigeot@wolfpond.org> wrote:
> > > I took advantage of an available dual-Xeon box to do some scalability
> > > benchmarking.
> > > 
> > > Some rapid conclusions:
> > > - NetBSD has trouble handling high numbers of database clients but
> > > still manages to scale when numbers are low
> > 
> > Would you be able to try using local UNIX socket?  NetBSD should have
> > nearly linear scalability, unless there is some regression.  Going
> > through 127.0.0.1 is what kills us, I believe.
> 
> You're right, it's much better with a Unix socket. NetBSD follows the
> FreeBSD curve up to 24 clients (including the peak).
> Sadly, it all goes downhill from there.

Good to know.  Do you have a graph somewhere?

> Performance with 32 clients is on par with the initial 127.0.0.1 results.
> Tests with 48 or more clients are still running and will probably yield
> bad results too.
> The system is almost completely unresponsive. Just seeing 'ls' appear on
> the remote shell I use takes about 20 seconds with a long pause between
> the two letters.

This is somewhat unexpected.  Do you use some unusual configuration or just
a plain GENERIC kernel?  Could you perhaps run pgbench under lockstat(8)?

Thank you.

-- 
Mindaugas