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 1CA25D1DB0A for ; Fri, 16 Apr 2004 04:24:34 -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 91072-04 for ; Fri, 16 Apr 2004 04:24:31 -0300 (ADT) Received: from cmailm1.svr.pol.co.uk (cmailm1.svr.pol.co.uk [195.92.193.18]) by svr1.postgresql.org (Postfix) with ESMTP id 5A209D1D2DC for ; Fri, 16 Apr 2004 04:24:32 -0300 (ADT) Received: from modem-1925.llama.dialup.pol.co.uk ([217.135.183.133] helo=stromboli) by cmailm1.svr.pol.co.uk with esmtp (Exim 4.14) id 1BENiC-0002hc-DM; Fri, 16 Apr 2004 08:24:32 +0100 From: "Simon Riggs" To: "Josh Berkus" , Subject: Re: PostgreSQL and Linux 2.6 kernel. Date: Fri, 16 Apr 2004 08:24:46 +0100 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2911.0) In-reply-to: <200404151339.45615.josh@agliodbs.com> X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6700 Importance: Normal X-Virus-Scanned: by amavisd-new at postgresql.org X-Archive-Number: 200404/201 X-Sequence-Number: 6501 >Josh Berkus > > Treating the optimizer as a black box is something I'm very > used to from > > other RDBMS. My question is, how can you explicitly > re-write a query now > > to "improve" it? If there's no way of manipulating queries without > > actually re-writing the optimizer, we're now in a position where we > > aren't able to diagnose when the optimizer isn't working > effectively. > > Well, there is ... all of the various query cost parameters. They are very blunt instruments for such a delicate task. Surely someone of your experience might have benefit from something more? My feeling is, I would, though I want those tools as *a developer* rather than for tuning specific queries for people, which is always so sensitive to upgrades etc. > But, ultimately, improvements on the planner are still > bottlenecked by having > only one developer actually hacking the changes. > Do we have a clear list of optimizations we'd like to be working on? The TODO items aren't very related to specific optimizations... The only ones I was aware of was deferred subselect evaluation for DBT-3. ...sounds like there's more to discuss here, so I'll duck out now and get back to my current project... Best Regards, Simon Riggs