X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (av.hub.org [200.46.204.144]) by postgresql.org (Postfix) with ESMTP id F0E539DC84E for ; Tue, 17 Jan 2006 13:25:13 -0400 (AST) Received: from postgresql.org ([200.46.204.71]) by localhost (av.hub.org [200.46.204.144]) (amavisd-new, port 10024) with ESMTP id 13446-01 for ; Tue, 17 Jan 2006 13:25:10 -0400 (AST) X-Greylist: from auto-whitelisted by SQLgrey- Received: from mail.logix-tt.com (serval.logix-tt.com [213.239.221.42]) by postgresql.org (Postfix) with ESMTP id 321359DC81E for ; Tue, 17 Jan 2006 13:25:08 -0400 (AST) Received: from kingfisher.intern.logix-tt.com (Mc7a6.m.pppool.de [89.49.199.166]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by mail.logix-tt.com (Postfix) with ESMTP id 59A2D24400F; Tue, 17 Jan 2006 18:31:27 +0100 (CET) Received: from [127.0.0.1] (localhost [127.0.0.1]) by kingfisher.intern.logix-tt.com (Postfix) with ESMTP id 9B0B518148BBE; Tue, 17 Jan 2006 18:25:06 +0100 (CET) Message-ID: <43CD2872.40509@logix-tt.com> Date: Tue, 17 Jan 2006 18:25:06 +0100 From: Markus Schaber Organization: Logical Tracking and Tracing International AG, Switzerland User-Agent: Debian Thunderbird 1.0.7 (X11/20051017) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Michael Riess Cc: pgsql-performance@postgresql.org Subject: Re: Autovacuum / full vacuum References: <43CCC711.504@familyhealth.com.au> <43CCFA7E.1060307@zeut.net> In-Reply-To: X-Enigmail-Version: 0.93.0.0 Content-Type: text/plain; charset=ISO-8859-15 Content-Transfer-Encoding: 7bit X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, score=0.125 required=5 tests=[AWL=0.125] X-Spam-Score: 0.125 X-Spam-Level: X-Archive-Number: 200601/250 X-Sequence-Number: 16728 Hi, Michael, Michael Riess wrote: > But actually I never understood why the database system slows down at > all when there is much unused space in the files. Are the unused pages > cached by the system, or is there another reason for the impact on the > performance? No, they are not cached as such, but PostgreSQL caches whole pages, and you don't have only empty pages, but also lots of partially empty pages, so the signal/noise ratio is worse (means PostgreSQL has to fetch more pages to get the same data). Sequential scans etc. are also slower. And some file systems get slower when files get bigger or there are more files, but this effect should not really be noticeable here. HTH, Markus -- Markus Schaber | Logical Tracking&Tracing International AG Dipl. Inf. | Software Development GIS Fight against software patents in EU! www.ffii.org www.nosoftwarepatents.org