Re: possible improvement between G4 and G5

From: Josh Berkus <josh(at)agliodbs(dot)com>
To: "Aaron Werman" <awerman2(at)hotmail(dot)com>, "Qing Zhao" <qzhao(at)quotefx(dot)net>, "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: <pgsql-performance(at)postgresql(dot)org>
Subject: Re: possible improvement between G4 and G5
Date: 2004-04-06 18:52:17
Message-ID: 200404061152.17997.josh@agliodbs.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Aaron,

> I'm surprised by this thought. I tend to hit CPU bottlenecks more often than
> I/O ones. In most applications, db I/O is a combination of buffer misses and
> logging, which are both reasonably constrained.

Not my experience at all. In fact, the only times I've seen modern platforms
max out the CPU was when:
a) I had bad queries with bad plans, or
b) I had reporting queires that did a lot of calculation for display (think
OLAP).

Otherwise, on the numerous servers I administrate, RAM spikes, and I/O
bottlenecks, but the CPU stays almost flat.

Of course, most of my apps are large databases (i.e. too big for RAM) with a
heavy transaction-processing component.

What kind of applications are you running?

--
-Josh Berkus
Aglio Database Solutions
San Francisco

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Aaron Werman 2004-04-06 20:22:46 Re: possible improvement between G4 and G5
Previous Message Stefan Kaltenbrunner 2004-04-06 16:22:49 Re: good pc but bad performance,why?