Received: from localhost (maia-1.hub.org [200.46.204.191]) by postgresql.org (Postfix) with ESMTP id 37ED69FB876 for ; Fri, 23 Feb 2007 13:37:58 -0400 (AST) Received: from postgresql.org ([200.46.204.71]) by localhost (mx1.hub.org [200.46.204.191]) (amavisd-new, port 10024) with ESMTP id 53216-01 for ; Fri, 23 Feb 2007 13:37:47 -0400 (AST) X-Greylist: from auto-whitelisted by SQLgrey-1.7.4 Received: from noel.decibel.org (noel.decibel.org [67.100.216.10]) by postgresql.org (Postfix) with ESMTP id A9C7D9FB7AC for ; Fri, 23 Feb 2007 13:37:47 -0400 (AST) Received: by noel.decibel.org (Postfix, from userid 1001) id DCF9E5648A; Fri, 23 Feb 2007 11:37:39 -0600 (CST) Received: (hashcash-sendmail, from uid 1001); Fri, 23 Feb 2007 11:37:36 -0600 Date: Fri, 23 Feb 2007 11:37:36 -0600 From: "Jim C. Nasby" To: Csaba Nagy Cc: Mark Stosberg , postgres performance list Subject: Re: Using the 8.2 autovacuum values with 8.1 Message-ID: <20070223173736.GA19527@nasby.net> References: <20070221185544.GX25424@alvh.no-ip.org> <45DCAD50.8080104@zeut.net> <1172222011.3772.13.camel@coppola.muc.ecircle.de> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1172222011.3772.13.camel@coppola.muc.ecircle.de> X-Operating-System: FreeBSD 6.1-RELEASE-p10 amd64 X-Distributed: Join the Effort! http://www.distributed.net User-Agent: Mutt/1.5.13 (2006-08-11) X-Hashcash: 1:20:070223:nagy@ecircle-ag.com::shtti5iNbjG/uliH:000000000000000000 0000000000000000000000002mKR X-Hashcash: 1:20:070223:mark@summersault.com::P13Fmt/iDxJs4Wzm:00000000000000000 000000000000000000000000Aw0+ X-Hashcash: 1:20:070223:pgsql-performance@postgresql.org::RgtqhHZt3o9AHBhY:00000 0000000000000000000000003gEd X-Virus-Scanned: Maia Mailguard 1.0.1 X-Archive-Number: 200702/247 X-Sequence-Number: 23301 On Fri, Feb 23, 2007 at 10:13:31AM +0100, Csaba Nagy wrote: > You likely don't need the nightly full vacuum run... we also do here a > nightly vacuum beside autovacuum, but not a full one, only for tables > which are big enough that we don't want autovacuum to touch them in high > business time but they have enough change that we want a vacuum on them > frequent enough. I discover them by checking the stats, for example: Something else I like doing is a periodic vacuumdb -av and capture the output. It's a good way to keep an eye on FSM utilization. Once you've got vacuuming under control you can probably just do that once a month or so. -- Jim Nasby jim@nasby.net EnterpriseDB http://enterprisedb.com 512.569.9461 (cell)