From owner-freebsd-questions@FreeBSD.ORG Fri Jan 2 13:05:12 2004 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 3254016A4CE for ; Fri, 2 Jan 2004 13:05:12 -0800 (PST) Received: from smithers.neuro.mcw.edu (smithers.neuro.mcw.edu [141.106.106.66]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5FE9C43D48 for ; Fri, 2 Jan 2004 13:05:10 -0800 (PST) (envelope-from bacon@smithers.neuro.mcw.edu) Received: from smithers.neuro.mcw.edu (localhost [127.0.0.1]) i02L51Xg040359; Fri, 2 Jan 2004 15:05:01 -0600 (CST) (envelope-from bacon@smithers.neuro.mcw.edu) Received: from localhost (localhost [[UNIX: localhost]]) by smithers.neuro.mcw.edu (8.12.9/8.12.9/Submit) id i02L4xsM040355; Fri, 2 Jan 2004 15:04:59 -0600 (CST) From: Jason Bacon Organization: MCW To: Francisco Reyes Date: Fri, 2 Jan 2004 15:04:58 -0600 User-Agent: KMail/1.5.2 References: <3FF31E4B.1070305@edgefocus.com> <200401021113.46323.jbacon@mcw.edu> <20040102131546.B72627@zoraida.natserv.net> In-Reply-To: <20040102131546.B72627@zoraida.natserv.net> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200401021504.58573.jbacon@mcw.edu> cc: Scott Mitchell cc: Sean Hafeez cc: freebsd-questions@freebsd.org Subject: Re: What do you use? X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 02 Jan 2004 21:05:12 -0000 On Friday 02 January 2004 07:19 am, Francisco Reyes wrote: > On Fri, 2 Jan 2004, Jason Bacon wrote: > > Well, I'm in a position to provide some comparison data with not-too-many > > variables, > > Thanks much for sharing the results. > > They seem close enough that someone who is price concious or on a limited > budget may want to consider the 3Ware. > > However, what a test like that may not show well is the difference when > having multi-user access. I think on that case the SCSI disks may do > better. Actually, I think the test *does* show that. It clearly showed that the SCSI system was somewhat less impacted by RAID access. Hence it would follow that the SCSI system could handle a somewhat heavier load before seeing a performance drop. The choice would depend how much the extra bandwidth is worth to you, assuming your usage patterns could make use of it at times. I can tell you that our users have never complained about performance of the server. We have relatively few users (1-10 at a time) running fairly intense data analysis (on the order of 1 hour total CPU time + processing gigabytes of 3d brain images ) either running on the server or over NFS. If you're wondering how they perform under heavy process loads, note that Yahoo is using the 3ware controllers extensively. Maybe they'd be willing to share their experience... Regards, Jason