From owner-freebsd-scsi@FreeBSD.ORG Wed May 21 00:36:57 2008 Return-Path: Delivered-To: freebsd-scsi@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id A6419106564A for ; Wed, 21 May 2008 00:36:57 +0000 (UTC) (envelope-from scottl@samsco.org) Received: from pooker.samsco.org (pooker.samsco.org [168.103.85.57]) by mx1.freebsd.org (Postfix) with ESMTP id 644B78FC1F for ; Wed, 21 May 2008 00:36:57 +0000 (UTC) (envelope-from scottl@samsco.org) Received: from phobos.local ([192.168.254.200]) (authenticated bits=0) by pooker.samsco.org (8.13.8/8.13.8) with ESMTP id m4L0amHO036386; Tue, 20 May 2008 18:36:49 -0600 (MDT) (envelope-from scottl@samsco.org) Message-ID: <48336EA0.3050109@samsco.org> Date: Tue, 20 May 2008 18:36:48 -0600 From: Scott Long User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US; rv:1.8.1.13) Gecko/20080313 SeaMonkey/1.1.9 MIME-Version: 1.0 To: Warren Guy References: <4832C397.3090004@calorieking.com> <4832E0EE.3030402@samsco.org> <4832E6C2.7040205@calorieking.com> In-Reply-To: <4832E6C2.7040205@calorieking.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Spam-Status: No, score=-4.2 required=3.8 tests=ALL_TRUSTED,AWL,BAYES_00 autolearn=ham version=3.1.8 X-Spam-Checker-Version: SpamAssassin 3.1.8 (2007-02-13) on pooker.samsco.org Cc: freebsd-scsi@freebsd.org Subject: Re: Very poor performance from Dell/LSI Logic SAS 3000 series SATA/SAS RAID controller FreeBSD 6.3 X-BeenThere: freebsd-scsi@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: SCSI subsystem List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 21 May 2008 00:36:57 -0000 Warren Guy wrote: > Scott, > > Thanks a lot for that. This seems to have alleviated the problem, I'm seeing > decent performance now in my limited benchmark. It seems quite odd to me that > the write cache is not enabled by default, but oh well. > > Thanks again for your help! > > Warren For data reliability, you really don't want it enabled by default. The problem is that SATA/ATA performs so poorly without it that everyone turns it on and lives with the consequences. The tweak that I recommended puts it in line with what the FreeBSD ATA driver has been doing for years. According to your original benchmark, Linux performs better on the sequential tests, but those simply aren't representative of most people's workloads. Linux indeed has some tricks to make sequential benchmarks perform well, but they aren't tricks that I'm all that interested in implementing in FreeBSD (though increasing the maxio size for 64-bit platforms would help and has few detrimental effects). The same benchmark shows that FreeBSD performs just as well, if not better, than Linux in random tests, even without the write cache enabled. Those tests are more representative of typical workloads. So, it's up to you to analyze what kind of workload you expect, and make the appropriate tradeoffs. Scott