From owner-freebsd-fs@FreeBSD.ORG Tue Nov 13 02:44:31 2012 Return-Path: Delivered-To: freebsd-fs@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id 24ECE3E5 for ; Tue, 13 Nov 2012 02:44:31 +0000 (UTC) (envelope-from bfriesen@simple.dallas.tx.us) Received: from blade.simplesystems.org (blade.simplesystems.org [65.66.246.74]) by mx1.freebsd.org (Postfix) with ESMTP id D03DC8FC08 for ; Tue, 13 Nov 2012 02:44:30 +0000 (UTC) Received: from freddy.simplesystems.org (freddy.simplesystems.org [65.66.246.65]) by blade.simplesystems.org (8.14.4+Sun/8.14.4) with ESMTP id qAD2Vn5O016177; Mon, 12 Nov 2012 20:31:49 -0600 (CST) Date: Mon, 12 Nov 2012 20:31:48 -0600 (CST) From: Bob Friesenhahn X-X-Sender: bfriesen@freddy.simplesystems.org To: Jason Keltz Subject: Re: RHEL to FreeBSD file server In-Reply-To: <50A130B7.4080604@cse.yorku.ca> Message-ID: References: <50A130B7.4080604@cse.yorku.ca> User-Agent: Alpine 2.01 (GSO 1266 2009-07-14) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; format=flowed; charset=US-ASCII X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.2.2 (blade.simplesystems.org [65.66.246.90]); Mon, 12 Nov 2012 20:31:49 -0600 (CST) Cc: freebsd-fs@freebsd.org X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.14 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 13 Nov 2012 02:44:31 -0000 I am not able to attest to the type of configuration you are building. It all sounds good to me and it would surely work fine with Solaris, and so likely with FreeBSD as well. The only concern might be that 64 GB of memory is not so much according to today's hardware standards. New hardware is easily able to fit 128 GB, 256 GB, or even 512 GB. If the working set from all those clients is bigger than will fit in 64 GB of RAM, then your disks will be working much harder than they should be. With zfs, disks don't do any reading (reads will stall) while zfs is writing a transaction group. A proper server should be doing mostly writing because key data is already in cache. With so many clients, make sure that your intent log FLASH devices are mirrored and be prepared to replace them periodically. Bob -- Bob Friesenhahn bfriesen@simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/