Received: from maia.hub.org (maia-2.hub.org [200.46.204.251]) by mail.postgresql.org (Postfix) with ESMTP id 1AEFFB5DBFF for ; Thu, 16 Jun 2011 11:18:44 -0300 (ADT) Received: from mail.postgresql.org ([200.46.204.86]) by maia.hub.org (mx1.hub.org [200.46.204.251]) (amavisd-maia, port 10024) with ESMTP id 62394-01 for ; Thu, 16 Jun 2011 14:18:37 +0000 (UTC) X-Greylist: from auto-whitelisted by SQLgrey-1.7.6 Received: from extern.innogames.de (extern.innogames.de [80.252.99.240]) by mail.postgresql.org (Postfix) with ESMTP id EF6A2B5DBFA for ; Thu, 16 Jun 2011 11:18:37 -0300 (ADT) Received: from [172.16.5.29] (unknown [212.48.107.10]) by extern.innogames.de (Postfix) with ESMTPA id 1299F28E21B for ; Thu, 16 Jun 2011 16:18:37 +0200 (CEST) Subject: Re: Missing clogs and pg_upgrade. What to do without any clog backup? From: Bernhard Schrader To: pgsql-admin@postgresql.org In-Reply-To: <1308226904.2334.15.camel@schrader-laptop> References: <1308226904.2334.15.camel@schrader-laptop> Content-Type: text/plain; charset="UTF-8" Organization: Innogames GmbH Date: Thu, 16 Jun 2011 16:18:35 +0200 Message-ID: <1308233915.2334.47.camel@schrader-laptop> Mime-Version: 1.0 X-Mailer: Evolution 2.28.3 Content-Transfer-Encoding: 7bit X-Virus-Scanned: Maia Mailguard 1.0.1 X-Spam-Status: No, hits=-1.91 tagged_above=-10 required=5 tests=BAYES_00=-1.9, T_RP_MATCHES_RCVD=-0.01 X-Spam-Level: X-Archive-Number: 201106/96 X-Sequence-Number: 37722 Am Donnerstag, den 16.06.2011, 14:21 +0200 schrieb Bernhard Schrader: > Hi out there, > > i migrated some month ago our databases from 8.4 to 9.0 and now getting > errors like: > > 2011-06-12 06:25:33 CEST ERROR: could not access status of transaction > 59768467 > 2011-06-12 06:25:33 CEST DETAIL: Could not open file "pg_clog/0038": No > such file or directory. > 2011-06-12 06:25:33 CEST CONTEXT: automatic vacuum of table > "database.pg_toast.pg_toast_19029" > > Well, I googled so far and also saw that there are other known issues > and also a fix script. But in all cases a backup of the clogs would have > been necessary. I don't have these files. Is there any way to fix my > servers? > > I have migrated around 200 servers, until now 40 servers have the clog > issues. i will run the script on all other machines. But for the 40 > servers i need a solution. > > Would a normal dump & restore fix the issues? This would be the worst > way i could imagine but if there is no way around i will have to do it. > > > thanks in advance > > regards > Bernhard Schrader > > Hi again, i just tried the pg_upgrade_fix.sql on databases where the errors didnt appear yet, this is what i get partially: VACUUM FREEZE pg_toast.pg_toast_182129; psql:pg_upgrade_tmp.sql:43: ERROR: could not access status of transaction 252555499 DETAIL: Could not open file "pg_clog/00F0": No such file or directory. Well, right now i dont know how to handle this. Is it now clean or is it worse than before? Just to get it in my mind. The clog is just a commit log, and the clogs which are missing are just some transactions that where used some months ago, right? So, even if the clogs are missing there should be no data loss or am i totally wrong? Well a recovery without them could be difficult but if its not necessary at all why can't i just start from zero with the clogs? regards Bernhard