Received: from localhost (unknown [200.46.204.184]) by developer.postgresql.org (Postfix) with ESMTP id AADFF2E008B for ; Tue, 6 May 2008 15:30:33 -0300 (ADT) Received: from developer.postgresql.org ([200.46.204.71]) by localhost (mx1.hub.org [200.46.204.184]) (amavisd-maia, port 10024) with ESMTP id 49515-02-5 for ; Tue, 6 May 2008 15:30:27 -0300 (ADT) X-Greylist: from auto-whitelisted by SQLgrey-1.7.5 Received: from relay-2.mail.nethere.net (relay-2.mail.nethere.net [66.63.128.162]) by developer.postgresql.org (Postfix) with ESMTP id E5DB12E0086 for ; Tue, 6 May 2008 15:30:29 -0300 (ADT) Received: from scan-1.mail.nethere.net (scan-1.mail.nethere.net [66.63.128.132]) by relay-2.mail.nethere.net (Postfix) with ESMTP id 6B30012C9DA; Tue, 6 May 2008 11:30:28 -0700 (PDT) Received: from mta-2.mail.nethere.net by scan-1.mail.nethere.net with LMTP id H9vfEynDFDG5; Tue, 6 May 2008 11:30:28 -0700 (PDT) Received: from lisa-james-computer.local (wsip-68-224-174-233.sd.sd.cox.net [68.224.174.233]) by mta-2.mail.nethere.net (Postfix) with ESMTP id 17C5E130C39; Tue, 6 May 2008 11:30:27 -0700 (PDT) Message-ID: <4820A3C2.4000307@emolecules.com> Date: Tue, 06 May 2008 11:30:26 -0700 From: Craig James User-Agent: Thunderbird 2.0.0.14 (Macintosh/20080421) MIME-Version: 1.0 To: Greg Smith CC: pgsql-performance@postgresql.org Subject: Re: RAID 10 Benchmark with different I/O schedulers References: <47FFA7FF.10801@postnewspapers.com.au> <544D07A2-EB3F-4D91-A749-0C4D5D38AB80@gmail.com> <481F992D.7070509@emolecules.com> In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Virus-Scanned: Maia Mailguard 1.0.1 X-Archive-Number: 200805/104 X-Sequence-Number: 30072 Greg Smith wrote: > On Mon, 5 May 2008, Craig James wrote: > >> pgbench -i -s 20 -U test > > That's way too low to expect you'll see a difference in I/O schedulers. > A scale of 20 is giving you a 320MB database, you can fit the whole > thing in RAM and almost all of it on your controller cache. What's > there to schedule? You're just moving between buffers that are > generally large enough to hold most of what they need. Test repeated with: autovacuum enabled database destroyed and recreated between runs pgbench -i -s 600 ... pgbench -c 10 -t 50000 -n ... I/O Sched AVG Test1 Test2 --------- ----- ----- ----- cfq 705 695 715 noop 758 769 747 deadline 741 705 775 anticipatory 494 477 511 I only did two runs of each, which took about 24 minutes. Like the first round of tests, the "noise" in the measurements (about 10%) exceeds the difference between scheduler-algorithm performance, except that "anticipatory" seems to be measurably slower. So it still looks like cfq, noop and deadline are more or less equivalent when used with a battery-backed RAID. Craig