Received: from localhost (unknown [200.46.204.183]) by postgresql.org (Postfix) with ESMTP id 58ED9650177 for ; Thu, 21 Aug 2008 02:34:03 -0300 (ADT) Received: from postgresql.org ([200.46.204.86]) by localhost (mx1.hub.org [200.46.204.183]) (amavisd-maia, port 10024) with ESMTP id 57235-02-5 for ; Thu, 21 Aug 2008 02:33:56 -0300 (ADT) X-Greylist: domain auto-whitelisted by SQLgrey-1.7.6 Received: from rv-out-0506.google.com (rv-out-0506.google.com [209.85.198.237]) by postgresql.org (Postfix) with ESMTP id 2BEE865015E for ; Thu, 21 Aug 2008 02:30:39 -0300 (ADT) Received: by rv-out-0506.google.com with SMTP id b25so1270970rvf.43 for ; Wed, 20 Aug 2008 22:30:38 -0700 (PDT) Received: by 10.141.105.18 with SMTP id h18mr475166rvm.207.1219290065121; Wed, 20 Aug 2008 20:41:05 -0700 (PDT) Received: by 10.141.197.1 with HTTP; Wed, 20 Aug 2008 20:41:05 -0700 (PDT) Message-ID: Date: Wed, 20 Aug 2008 21:41:05 -0600 From: "Scott Marlowe" To: "Christiaan Willemsen" Subject: Re: How to setup disk spindles for best performance Cc: pgsql-performance@postgresql.org In-Reply-To: <972BDF1A-B4ED-46EC-B416-0CDF15A0439A@technocon.com> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline References: <972BDF1A-B4ED-46EC-B416-0CDF15A0439A@technocon.com> X-Virus-Scanned: Maia Mailguard 1.0.1 X-Spam-Status: No, hits=0 tagged_above=0 required=5 tests=none X-Spam-Level: X-Archive-Number: 200808/237 X-Sequence-Number: 31118 On Wed, Aug 20, 2008 at 4:25 PM, Christiaan Willemsen wrote: > I'm currently trying to find out what the best configuration is for our new > database server. It will server a database of about 80 GB and growing fast. > The new machine has plenty of memory (64GB) and 16 SAS disks, of wich two > are already in use as a mirror for the OS. Got almost the same setup with 32 Gig of ram... I'm running the Areca 1680 controller with 512M battery backed cache and I'm quite happy with the performance so far. Mine is set up to use the OS mirror set for the pg_xlog as well, but on a separate logical partition. I figure the RAID controller will even out the writes there to keep it fast. The next 12 drives are a RAID-10 set and the last two are hot spares. On an array this big you should always have at least one hot spare. Generally breaking the disks up for index versus tables etc is a lot of work for minimal gain. A good RAID controller will make up for having to do that, and usually do better, since if the indexes are getting hit a LOT then you have all 12 disks in the RAID-10 working together. But what's your workload look like? Lotsa updates, inserts, deletes, big selects, bulk loads in the middle of the day, etc...?