Received: from localhost (wm.hub.org [200.46.204.128]) by postgresql.org (Postfix) with ESMTP id 13DC29FB259 for ; Thu, 26 Oct 2006 06:22:19 -0300 (ADT) Received: from postgresql.org ([200.46.204.71]) by localhost (mx1.hub.org [200.46.204.128]) (amavisd-new, port 10024) with ESMTP id 56575-01 for ; Thu, 26 Oct 2006 09:21:01 +0000 (UTC) X-Greylist: delayed 00:22:29.171777 by SQLgrey- Received: from svr4.postgresql.org (svr4.postgresql.org [66.98.251.159]) by postgresql.org (Postfix) with ESMTP id 7D8E99FB34F for ; Thu, 26 Oct 2006 06:21:00 -0300 (ADT) X-Greylist: domain auto-whitelisted by SQLgrey-1.7.3 Received: from pat.laterooms.com (fon.laterooms.com [194.24.251.1]) by svr4.postgresql.org (Postfix) with ESMTP id 2DAD65AF04F for ; Thu, 26 Oct 2006 08:58:30 +0000 (GMT) Received: from pat.laterooms.com (localhost.localdomain [127.0.0.1]) by pat.laterooms.com (Postfix) with ESMTP id 7B9D1C41C for ; Thu, 26 Oct 2006 09:58:27 +0100 (BST) Received: from teapot.laterooms.com (bum.net [194.24.251.10]) by pat.laterooms.com (Postfix) with SMTP id 5FD34C3E8 for ; Thu, 26 Oct 2006 09:58:27 +0100 (BST) Date: Thu, 26 Oct 2006 09:58:27 +0100 From: Gavin Hamill To: pgsql-performance@postgresql.org Subject: VACUUMs take twice as long across all nodes Message-Id: <20061026095827.9e85e1c0.gdh@laterooms.com> X-Mailer: Sylpheed version 2.2.4 (GTK+ 2.8.20; i486-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-AV-Checked: ClamAV using ClamSMTP X-Virus-Scanned: Maia Mailguard 1.0.1 X-Spam-Status: No, hits=1.277 tagged_above=0 required=5 tests=FORGED_RCVD_HELO, SPF_FAIL X-Spam-Level: * X-Archive-Number: 200610/544 X-Sequence-Number: 21673 Hullo, here's one of those dreadful touchy-feely hand-waving problems. Our 5-node 8.1.3 Slony system has just started taking /much/ longer to VACUUM ANALYZE.. The data set has not increased more than usual (nightly backups stand at 1.3GB, growing by 10MB per day), and no configuration has changed on the machines. Nodes 2 and 3 take only the tables necessary to run our search (10 out of the full 130) and are much lighter (only 7GB on disk cf. 30GB for the full master) , yet the nightly VACUUM FULL has jumped from 2 hours to 4 in the space of one day! Like I say, no config changes, no reboots / postmaster restarts, no extra processes, and every machine has a comfortable overhead of free page slots + relations.