X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [200.46.204.2]) by svr1.postgresql.org (Postfix) with ESMTP id 183E3D1E078 for ; Thu, 19 Feb 2004 18:06:22 +0000 (GMT) Received: from svr1.postgresql.org ([200.46.204.71]) by localhost (neptune.hub.org [200.46.204.2]) (amavisd-new, port 10024) with ESMTP id 59444-05 for ; Thu, 19 Feb 2004 14:06:21 -0400 (AST) Received: from easily.co.uk (mercury0.easily.co.uk [213.161.76.90]) by svr1.postgresql.org (Postfix) with ESMTP id 6DD0FD1DF93 for ; Thu, 19 Feb 2004 14:06:18 -0400 (AST) Received: from [213.152.63.90] (account f4vo5dsy5djd HELO chuckie.co.uk) by easily.co.uk (CommuniGate Pro SMTP 4.1.3) with ESMTP id 49684497; Thu, 19 Feb 2004 18:06:14 +0000 Message-ID: <4034FB11.7070305@chuckie.co.uk> Date: Thu, 19 Feb 2004 18:06:09 +0000 From: Nick Barr User-Agent: Mozilla Thunderbird 0.5 (Windows/20040207) X-Accept-Language: en-us, en MIME-Version: 1.0 To: josh@agliodbs.com Cc: Anjan Dave , pgsql-performance@postgresql.org Subject: Re: Postgresql on SAN References: <4BAFBB6B9CC46F41B2AD7D9F4BBAF78508C847@vt-pe2550-001.vantage.vantage.com> <200402190936.47208.josh@agliodbs.com> In-Reply-To: <200402190936.47208.josh@agliodbs.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Virus-Scanned: by amavisd-new at postgresql.org X-Archive-Number: 200402/227 X-Sequence-Number: 5751 Josh Berkus wrote: >Anjan, > > > >>Has anyone designed/implemented postgresql server on storage networks? >> >> > >Yes, Zapatec.com runs their stuff this way. Probably others as well. > > > >>Are there any design considerations? >> >> > >I don't know. Probably. > > > >>Are there any benchmarks for storage products (HBAs, Switches, Storage >>Arrays)? >> >> > >Not specific to PostgreSQL. I'm sure there are generic benchmarks. Keep >in mind that PostgreSQL needs lots of 2-way I/O, batch writes, and random >reads. > > > >>Any recommendation on the design, resources, references, keeping PG in >>mind? >> >> > >See above. Also keep in mind that PostgreSQL's use of I/O should improve >100% in version 7.5. > > > We run PG on a SAN array. We currently have it setup so a single PG instance runs off of a single LUN, this includes the WAL logs. Apart from that we have made no other special considerations; we just treat it as a fast RAID array. We haven't got to the stage where the speed of the SAN is a problem as load hasn't increased as expected. This will change, when it does I am sure the performance list will be hearing from us ;-). Out current limitations, as I see it, are amount of memory and then processing power. The only problem we have had was a dodgy set of kernel modules (drivers) for the fibre cards, this was because they were beta drivers and obviously still had a few bugs. This was solved by reverting to an older version. Everything has run smoothly since then (uptime is 153 days :-)). Nick