X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [200.46.204.144]) by svr1.postgresql.org (Postfix) with ESMTP id 0FB6632AC23 for ; Fri, 15 Oct 2004 00:55:54 +0100 (BST) Received: from svr1.postgresql.org ([200.46.204.71]) by localhost (av.hub.org [200.46.204.144]) (amavisd-new, port 10024) with ESMTP id 02518-10 for ; Thu, 14 Oct 2004 23:55:48 +0000 (GMT) Received: from davinci.ethosmedia.com (server226.ethosmedia.com [209.128.84.226]) by svr1.postgresql.org (Postfix) with ESMTP id A502932A9F8 for ; Fri, 15 Oct 2004 00:55:50 +0100 (BST) Received: from [64.81.245.111] (account josh@agliodbs.com HELO temoku.sf.agliodbs.com) by davinci.ethosmedia.com (CommuniGate Pro SMTP 4.1.8) with ESMTP id 6504705; Thu, 14 Oct 2004 16:57:14 -0700 From: Josh Berkus Reply-To: josh@agliodbs.com Organization: Aglio Database Solutions To: pgsql-performance@postgresql.org Subject: Re: First set of OSDL Shared Mem scalability results, some wierdness ... Date: Thu, 14 Oct 2004 16:57:45 -0700 User-Agent: KMail/1.6.2 Cc: "Simon Riggs" , References: In-Reply-To: MIME-Version: 1.0 Content-Disposition: inline Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Message-Id: <200410141657.45415.josh@agliodbs.com> X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, hits=0.0 tagged_above=0.0 required=5.0 tests= X-Spam-Level: X-Archive-Number: 200410/199 X-Sequence-Number: 8675 Simon, > If you draw a graph of speedup (y) against cache size as a > % of total database size, the graph looks like an upside-down "L" - i.e. > the graph rises steeply as you give it more memory, then turns sharply at a > particular point, after which it flattens out. The "turning point" is the > "sweet spot" we all seek - the optimum amount of cache memory to allocate - > but this spot depends upon the worklaod and database size, not on available > RAM on the system under test. Hmmm ... how do you explain, then the "camel hump" nature of the real performance? That is, when we allocated even a few MB more than the "optimum" ~190MB, overall performance stated to drop quickly. The result is that allocating 2x optimum RAM is nearly as bad as allocating too little (e.g. 8MB). The only explanation I've heard of this so far is that there is a significant loss of efficiency with larger caches. Or do you see the loss of 200MB out of 3500MB would actually affect the Kernel cache that much? Anyway, one test of your theory that I can run immediately is to run the exact same workload on a bigger, faster server and see if the desired quantity of shared_buffers is roughly the same. I'm hoping that you're wrong -- not because I don't find your argument persuasive, but because if you're right it leaves us without any reasonable ability to recommend shared_buffer settings. -- --Josh Josh Berkus Aglio Database Solutions San Francisco