From owner-freebsd-stable@FreeBSD.ORG Wed Mar 24 17:37:00 2010 Return-Path: Delivered-To: freebsd-stable@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id B61ED106566C for ; Wed, 24 Mar 2010 17:37:00 +0000 (UTC) (envelope-from michal@ionic.co.uk) Received: from mail1.sharescope.co.uk (pm1.ionic.co.uk [85.159.80.19]) by mx1.freebsd.org (Postfix) with ESMTP id 71A4F8FC0A for ; Wed, 24 Mar 2010 17:36:59 +0000 (UTC) Received: from localhost (unknown [127.0.0.1]) by mail1.sharescope.co.uk (Postfix) with ESMTP id AAEC5FC0AB for ; Wed, 24 Mar 2010 17:37:00 +0000 (UTC) X-Virus-Scanned: amavisd-new at sharescope.co.uk Received: from mail1.sharescope.co.uk ([127.0.0.1]) by localhost (mail1.sharescope.co.uk [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id Znd1F+PxxLAc for ; Wed, 24 Mar 2010 17:36:56 +0000 (GMT) Received: from [192.168.2.37] (office.ionic.co.uk [85.159.85.2]) (Authenticated sender: chris@sharescope.co.uk) by mail1.sharescope.co.uk (Postfix) with ESMTPSA id 4AF57FC0A2 for ; Wed, 24 Mar 2010 17:36:56 +0000 (GMT) Message-ID: <4BAA4DB6.80604@ionic.co.uk> Date: Wed, 24 Mar 2010 17:36:54 +0000 From: Michal User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.1.8) Gecko/20100227 Lightning/1.0b1 Thunderbird/3.0.3 MIME-Version: 1.0 To: freebsd-stable@freebsd.org References: <4BAA3409.6080406@ionic.co.uk> In-Reply-To: X-Enigmail-Version: 1.0.1 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Subject: Re: Multi node storage, ZFS X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 24 Mar 2010 17:37:00 -0000 On 24/03/2010 17:14, Freddie Cash wrote: > > Yes, that would be helpful (mirrored slogs, until we get slog removal > support). > > As would an L2ARC (cache) device in the head node. > > As well as lots and lots and lots of RAM. > > And as fast of ethernet NICs as you can get between the head node and the > storage nodes. > > And, and, and, and ... :) > As far as I know more RAM is more important the fast CPU, so RAM is the order of the day, and I guess it depends what you think fast CPU is, but I wasn't planning on a duel CPU or anything top of the range. I have some duel core's knocking around...I think testing will show how good/bad my calculations/assumptions are. Most are done in batched, nightly and weekly so extremely fast isn't THAT important as we are looking at device/server backup's and stored data which is moved off servers once a week. At the minute we are not looking at 100 user file system or anything along those lines. For NICS I can sort out a Gb switch or some point-to-point Gb connections betweeen the nodes, there is also the option is trying getting some cheap fibre cards (I have a few laying around) and a cheap fibre switch (something off ebay for testing might do) to have fibre between the nodes. This however all goes out the water for trying to do replication to other sites which are 100mb lines, but for the minute I will focus on 1 location to stop it getting too complex. There are quite a lot of hardware things which need to be done correctly, and yes I do need to look at lots of other things. But stage one is just getting a few Storage devices talking to a Storage controller and seeing if my ideas work (improve IO, improved redundancy, easy to add storage) Michael, I sort of understand what you are talking about with ZIL, but not completely, so thanks for the pointers, there are clearly things I have not thought about.