Received: from mail.perdue.net (c696732-a.west1.ia.home.com [24.22.227.157]) by postgresql.org (8.11.3/8.11.1) with ESMTP id f35ExJ838433 for ; Thu, 5 Apr 2001 10:59:20 -0400 (EDT) (envelope-from tperdue@mail.perdue.net) Received: (from tperdue@localhost) by mail.perdue.net (8.11.2/8.11.2/Debian 8.11.2-1) id f35Evll20285; Thu, 5 Apr 2001 09:57:47 -0500 Date: Thu, 5 Apr 2001 09:57:47 -0500 From: Tim Perdue To: Tom Lane Cc: pgsql-sql@postgresql.org Subject: Re: Memory and performance Message-ID: <20010405095747.L513@mail.perdue.net> References: <200104041715.f34HF8858323@postgresql.org> <14398.986435610@sss.pgh.pa.us> <20010405075250.H513@mail.perdue.net> <16726.986481967@sss.pgh.pa.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.3.12i In-Reply-To: <16726.986481967@sss.pgh.pa.us>; from tgl@sss.pgh.pa.us on Thu, Apr 05, 2001 at 10:46:07AM -0400 X-Archive-Number: 200104/82 X-Sequence-Number: 2422 On Thu, Apr 05, 2001 at 10:46:07AM -0400, Tom Lane wrote: > Tim Perdue writes: > > Our database is about 1GB in total size, the machine has 4GB, but the entire > > system is only using 1.2 GB, even during vacuum or a daily, extremely large > > query that requires a lot of grouping and sorting. > > What have you got the SortMem parameter (backend -S switch) set to? > That's about the only use I can think of for RAM beyond what's needed to > cache the whole database ... -i -o -F -S 32768 -B 2048 is our startup line. It appears to me that when you're vacuuming, it's physically moving all the bits around the disk. It doesn't read in the table, clean out the cruft, the flush it out. So I think this is where we're getting IO bound. We only have a 5x36 RAID - must not be sufficient. Tim -- Founder - PHPBuilder.com / Geocrawler.com Lead Developer - SourceForge VA Linux Systems