From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S267796AbUHER3M (ORCPT ); Thu, 5 Aug 2004 13:29:12 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S267797AbUHER3M (ORCPT ); Thu, 5 Aug 2004 13:29:12 -0400 Received: from holomorphy.com ([207.189.100.168]:63940 "EHLO holomorphy.com") by vger.kernel.org with ESMTP id S267796AbUHER24 (ORCPT ); Thu, 5 Aug 2004 13:28:56 -0400 Date: Thu, 5 Aug 2004 10:25:31 -0700 From: William Lee Irwin III To: "Mr. Berkley Shands" Cc: linux-kernel@vger.kernel.org Subject: Re: Severe I/O performance regression 2.6.6 to 2.6.7 or 2.6.8-rc3 Message-ID: <20040805172531.GC17188@holomorphy.com> Mail-Followup-To: William Lee Irwin III , "Mr. Berkley Shands" , linux-kernel@vger.kernel.org References: <41126811.7020607@dssimail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <41126811.7020607@dssimail.com> User-Agent: Mutt/1.5.6+20040722i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Aug 05, 2004 at 12:02:09PM -0500, Mr. Berkley Shands wrote: > Two severe disk read bugs: > In a nutshell (see attached for gory details). Moving from 2.6.6 to > 2.6.7 dropped multi-threaded RAID0 > read performance from 429MB/Sec to 81MB/Sec. Single threaded reads > improved 368MB/Sec to 418MB/Sec. > The code in drivers/md has no effect on this problem. Clearly this is a > thread access issue. Redhat ES3.0 > on x86_64 or i686. The underlying hardware is capable of 955MB/Sec disk > reads off 28 drives, > 541MB/Sec off 14 drives. Tuning I/O block size (11KB to 239KB) and > BLKRASET size (448 to 1024 or more) > helps a little. System idle goes from 0% to 50% (2.6.6 to 2.6.8-rc3). By any chance could you do binary search on the bk snapshots between 2.6.6 and 2.6.7? -- wli