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 97555D1B58E for ; Thu, 6 May 2004 13:01:56 -0300 (ADT) 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 32302-06 for ; Thu, 6 May 2004 13:01:35 -0300 (ADT) Received: from web13808.mail.yahoo.com (web13808.mail.yahoo.com [216.136.175.18]) by svr1.postgresql.org (Postfix) with SMTP id 139DDD1EAD1 for ; Thu, 6 May 2004 13:01:34 -0300 (ADT) Message-ID: <20040506160133.54511.qmail@web13808.mail.yahoo.com> Received: from [216.173.173.66] by web13808.mail.yahoo.com via HTTP; Thu, 06 May 2004 09:01:33 PDT Date: Thu, 6 May 2004 09:01:33 -0700 (PDT) From: Chris Gamache Reply-To: cgg007@yahoo.com Subject: Re: Postmaster hogs CPU To: "scott.marlowe" Cc: pgsql-admin@postgresql.org In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Virus-Scanned: by amavisd-new at postgresql.org X-Spam-Status: No, hits=0.9 tagged_above=0.0 required=5.0 tests=FROM_ENDS_IN_NUMS X-Spam-Level: X-Archive-Number: 200405/60 X-Sequence-Number: 13376 --- "scott.marlowe" wrote: > On Wed, 5 May 2004, Chris Gamache wrote: > Have you read: > > http://www.varlena.com/varlena/GeneralBits/Tidbits/perf.html > > yet? It may just be that you don't have enough shared_buffers or whatnot. Absolutely. I have a copy on my desk, and one under my pillow at night. :) PostgreSQL profiling is a bit of an art. I've been a settings-tweaker since I can remember. I feel like my config is pretty well balanced. Its that giant, oddball-query that throws the proverbial wrench in the gears. The guide is quick to warn that "too much" in any particular setting can be as bad as "not enough" ... That's not to say that someone else, with more expertise than I, couldn't look at what I have as far as settings and choose some better ones. The query_priority idea sounds like a promising idea. For queries that you /know/ will be expensive (or that you know you want to have run lickety-split), it'd do the trick. The next logical step would be an "intelligent" query priority engine. However, that might require some ESP (the X-Files kind) ... How do the other big RDBMS's do it? CG __________________________________ Do you Yahoo!? Win a $20,000 Career Makeover at Yahoo! HotJobs http://hotjobs.sweepstakes.yahoo.com/careermakeover