X-Original-To: pgsql-admin-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [200.46.204.2]) by svr1.postgresql.org (Postfix) with ESMTP id 0D71BD1D153 for ; Mon, 22 Mar 2004 16:40:18 +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 83612-07 for ; Mon, 22 Mar 2004 12:40:25 -0400 (AST) Received: from fep04-mail.bloor.is.net.cable.rogers.com (fep04-mail.bloor.is.net.cable.rogers.com [66.185.86.74]) by svr1.postgresql.org (Postfix) with ESMTP id 7F6EED1C9B5 for ; Mon, 22 Mar 2004 12:40:14 -0400 (AST) Received: from phlogiston.dydns.org ([65.49.121.67]) by fep04-mail.bloor.is.net.cable.rogers.com (InterMail vM.5.01.05.12 201-253-122-126-112-20020820) with ESMTP id <20040322163943.NBZS323940.fep04-mail.bloor.is.net.cable.rogers.com@phlogiston.dydns.org> for ; Mon, 22 Mar 2004 11:39:43 -0500 Received: by phlogiston.dydns.org (Postfix, from userid 1000) id 82D8E3C0E; Mon, 22 Mar 2004 11:40:24 -0500 (EST) Date: Mon, 22 Mar 2004 11:40:24 -0500 From: Andrew Sullivan To: pgsql-admin@postgresql.org Subject: Re: Hardware and OS for postgresql Message-ID: <20040322164024.GD11738@phlogiston.dyndns.org> Mail-Followup-To: pgsql-admin@postgresql.org References: <2f57764a.0403201952.5df3b33b@posting.google.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <2f57764a.0403201952.5df3b33b@posting.google.com> User-Agent: Mutt/1.5.5.1+cvs20040105i X-Authentication-Info: Submitted using SMTP AUTH LOGIN at fep04-mail.bloor.is.net.cable.rogers.com from [65.49.121.67] using ID at Mon, 22 Mar 2004 11:39:43 -0500 X-Virus-Scanned: by amavisd-new at postgresql.org X-Archive-Number: 200403/287 X-Sequence-Number: 12856 On Sat, Mar 20, 2004 at 07:52:08PM -0800, Ghazan Haider wrote: > (1) Scaling in which direction will help postgresql best, given the > queries are CPU, memory, io and disk-intensive? I understand dual-CPUs > will help in certain circumstances, but say for large subqueries which > are built in the memory, will it help to have a server in which there > are memory segments dedicated to each CPUs? In this regard, does > anyone know of a specific Sun Ultra, RS/6000 HP or Dell server whose > architecture is especially favorable to Postgresql or databases in > general? I think dedicating memory to individual CPUs is likely to be counter-productive. You won't be able to share the memory for filesystem buffers, which are the things most likely to give you the biggest return. I _can_ tell you that I've recently had an opportunity to work with a fairly large IBM RS/6000 (P650), and it is one fast machine. It's sort of unfair to compare it to UltraSPARC II machines, but I certainly am impressed as compared to Sun's E4500. A -- Andrew Sullivan | ajs@crankycanuck.ca In the future this spectacle of the middle classes shocking the avant- garde will probably become the textbook definition of Postmodernism. --Brad Holland