X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (av.hub.org [200.46.204.144]) by svr1.postgresql.org (Postfix) with ESMTP id 2EF52D8527 for ; Wed, 31 Aug 2005 10:51:42 -0300 (ADT) Received: from svr1.postgresql.org ([200.46.204.71]) by localhost (av.hub.org [200.46.204.144]) (amavisd-new, port 10024) with ESMTP id 46809-09 for ; Wed, 31 Aug 2005 13:51:40 +0000 (GMT) Received: from mail.libertyrms.com (unknown [207.219.45.62]) by svr1.postgresql.org (Postfix) with ESMTP id 872D6D8509 for ; Wed, 31 Aug 2005 10:51:38 -0300 (ADT) Received: from dba5.int.libertyrms.com ([10.1.3.44]) by mail.libertyrms.com with esmtp (Exim 4.22) id 1EAT0B-0005xV-Ej for pgsql-performance@postgresql.org; Wed, 31 Aug 2005 09:51:43 -0400 Message-ID: <4315B5EF.5020509@ca.afilias.info> Date: Wed, 31 Aug 2005 09:51:43 -0400 From: Brad Nicholson User-Agent: Mozilla Thunderbird 0.8 (X11/20040913) X-Accept-Language: en-us, en MIME-Version: 1.0 To: pgsql-performance@postgresql.org Subject: Re: 'Real' auto vacuum? References: <4314DBE0.9070708@telogis.com> <200508301558.34697.josh@agliodbs.com> <00f601c5adeb$56fcb910$f20214ac@bite.lt> In-Reply-To: <00f601c5adeb$56fcb910$f20214ac@bite.lt> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-SA-Exim-Mail-From: bnichols@ca.afilias.info X-SA-Exim-Scanned: No; SAEximRunCond expanded to false X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, hits=0.311 required=5 tests=[AWL=-0.220, FORGED_RCVD_HELO=0.05, INFO_TLD=0.481] X-Spam-Level: X-Archive-Number: 200508/556 X-Sequence-Number: 14303 Mindaugas Riauba wrote: >>>When a row is orphaned it's added to a list of possibly available rows. >>>When a new row is needed the list of possible rows is examined and the >>>first one with a transaction id less then the lowest running transaction >>>id is chosen to be the new row? These rows can be in a heap so it's >>>really fast to find one. >>> >>> >>This is the long-term plan. However, it's actually a lot harder than it >>sounds. Patches welcome. >> >> > > Some ETA? Since that would be the most welcome addition for us. We >have few very heavily updated databases where table bloat and constant >vacuuming is killing performance. > > > How often are you vacuuming (the definition of 'constantly' tends to vary)? Are you vacuuming the whole database each time? If so, identify which tables are being updated frequently, and vacuum those often. Vacuum other tables less frequently. Also, are you you using VACUUM FULL (if so, you certainly don't want to be). -- Brad Nicholson 416-673-4106 bnichols@ca.afilias.info Database Administrator, Afilias Canada Corp.