From owner-freebsd-questions@FreeBSD.ORG Thu Nov 30 20:01:55 2006 Return-Path: X-Original-To: freebsd-questions@freebsd.org Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id 4350C16A5B6 for ; Thu, 30 Nov 2006 20:01:55 +0000 (UTC) (envelope-from cswiger@mac.com) Received: from smtpout.mac.com (smtpout.mac.com [17.250.248.172]) by mx1.FreeBSD.org (Postfix) with ESMTP id B2CD443DE2 for ; Thu, 30 Nov 2006 19:57:09 +0000 (GMT) (envelope-from cswiger@mac.com) Received: from mac.com (smtpin05-en2 [10.13.10.150]) by smtpout.mac.com (Xserve/8.12.11/smtpout02/MantshX 4.0) with ESMTP id kAUJv8cG013629; Thu, 30 Nov 2006 11:57:08 -0800 (PST) Received: from [17.214.13.96] (a17-214-13-96.apple.com [17.214.13.96]) (authenticated bits=0) by mac.com (Xserve/smtpin05/MantshX 4.0) with ESMTP id kAUJv46Q021673; Thu, 30 Nov 2006 11:57:07 -0800 (PST) In-Reply-To: <200611301942.kAUJgMiv012412@mac.com> References: <200611301942.kAUJgMiv012412@mac.com> Mime-Version: 1.0 (Apple Message framework v752.2) X-Priority: 3 Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Message-Id: Content-Transfer-Encoding: 7bit From: Chuck Swiger Date: Thu, 30 Nov 2006 11:57:03 -0800 To: Derrick MacPherson X-Mailer: Apple Mail (2.752.2) X-Brightmail-Tracker: AAAAAA== X-Brightmail-scanned: yes Cc: freebsd-questions@freebsd.org Subject: Re: degraded RAID performance after OS upgrade, and drives added. X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 30 Nov 2006 20:01:55 -0000 On Nov 30, 2006, at 11:40 AM, Derrick MacPherson wrote: > That seems like a pretty crazy drop in performance, more than one > would expect. The machine is busy but not busy enough to warrant > this.. Imo.. Is there a way to test to confirm? Using dd is a trivial benchmark, and not especially precise but good enough to give rough answers; otherwise, there are lots of I/O benchmarks in ports like iozone or even "diskinfo -t". Try using different block sizes while reading and writing with dd, and you'll probably find some useful info. While it is likely that you can adjust the RAID-5 stripe size, change the write-caching from write-thru to write-back (if your RAID has a battery, anyway, otherwise this is dangerous), etc, using a 3-disk RAID-5 volume is just not a great idea-- RAID-5 is happier with more drives than the minimum of 3 to get better parallelism and reduce the overhead for parity info. -- -Chuck