Received: from oss.sgi.com (localhost [127.0.0.1]) by oss.sgi.com (8.12.3/8.12.3) with ESMTP id g3PKInwJ024456 for ; Thu, 25 Apr 2002 13:18:49 -0700 Received: (from majordomo@localhost) by oss.sgi.com (8.12.3/8.12.3/Submit) id g3PKIn3w024455 for linux-xfs-outgoing; Thu, 25 Apr 2002 13:18:49 -0700 X-Authentication-Warning: oss.sgi.com: majordomo set sender to owner-linux-xfs@oss.sgi.com using -f Received: from zeus-e8.americas.sgi.com ([198.149.7.103]) by oss.sgi.com (8.12.3/8.12.3) with SMTP id g3PKIfwJ024424 for ; Thu, 25 Apr 2002 13:18:41 -0700 Received: from daisy-e185.americas.sgi.com (daisy-e194.americas.sgi.com [128.162.194.214]) by zeus-e8.americas.sgi.com (SGI-SGI-8.9.3/americas-smart-nospam1.1) with ESMTP id PAA43294; Thu, 25 Apr 2002 15:19:06 -0500 (CDT) Received: from jen.americas.sgi.com (jen.americas.sgi.com [128.162.187.49]) by daisy-e185.americas.sgi.com (SGI-8.9.3/SGI-server-1.7) with ESMTP id PAA55203; Thu, 25 Apr 2002 15:19:06 -0500 (CDT) Received: by jen.americas.sgi.com (8.11.6/SGI-client-1.7) id g3PKJ5O14302; Thu, 25 Apr 2002 15:19:05 -0500 Subject: Re: poor io performance with xfs+raid5 From: Steve Lord To: Mike Eldridge Cc: Simon Matter , Justin Coffey , linux-xfs@oss.sgi.com In-Reply-To: <20020425144025.N16048@ornery.cafes.net> References: <3CC85999.501431E5@ch.sauter-bc.com> <20020425144025.N16048@ornery.cafes.net> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.3 Date: 25 Apr 2002 15:19:05 -0500 Message-Id: <1019765945.12905.102.camel@jen.americas.sgi.com> Mime-Version: 1.0 Sender: owner-linux-xfs@oss.sgi.com Precedence: bulk On Thu, 2002-04-25 at 14:40, Mike Eldridge wrote: > On Thu, Apr 25, 2002 at 09:31:37PM +0200, Simon Matter wrote: > > > > Hi, > > > > Hardware RAID is not affected at all. All levels of Software RAID beside > > RAID5 are okay. The only problem is with Software RAID5 and internal > > log. Under some circumstances, this config can slow down performance > > _dramatically_. > > it should be noted that i am using hardware raid5: > > > > > > the system is configured as such: > > > > > 60GB RAID5 (escalade 7850 + 4x20GB IBM deskstar) mounted on /var > > > > > log=internal,bsize=4096,blocks=1839 > > > > > agcount=58,agsize=262144 blocks > > > > > realtime=none > > via a 3ware escalade 7850 ata-raid card. > Well, Duh! I should have seen that first time around, I get into the habit of reading my email too fast! We may be able to fix some things, if we can remake the filesystem. First you need to know the stripe unit of your raid - we can feed this into XFS to make it do stripe aligned allocations. This has to be done by hand on linux. Take a look at the mkfs.xfs man page and the section on sunit and swidth options. Probably bump your log size up from the default somewhat, not sure how it ended up as 1839 that is scary. You still have not said which kernel version you are running beyond 2.4, unless I speed read over that too. Steve -- Steve Lord voice: +1-651-683-3511 Principal Engineer, Filesystem Software email: lord@sgi.com