X-Original-To: pgsql-admin-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [200.46.204.144]) by svr1.postgresql.org (Postfix) with ESMTP id 34EEB3A5A87 for ; Tue, 11 Jan 2005 20:00:14 +0000 (GMT) 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 71588-03 for ; Tue, 11 Jan 2005 19:59:57 +0000 (GMT) Received: from sss.pgh.pa.us (sss.pgh.pa.us [66.207.139.130]) by svr1.postgresql.org (Postfix) with ESMTP id 6E1833A5A81 for ; Tue, 11 Jan 2005 19:59:57 +0000 (GMT) Received: from sss2.sss.pgh.pa.us (tgl@localhost [127.0.0.1]) by sss.pgh.pa.us (8.13.1/8.13.1) with ESMTP id j0BJxt3b003616; Tue, 11 Jan 2005 14:59:56 -0500 (EST) To: "Lee Wu" Cc: pgsql-admin@postgresql.org Subject: Re: Vacuum full - disk space eaten by WAL logfiles In-reply-to: References: Comments: In-reply-to "Lee Wu" message dated "Tue, 11 Jan 2005 11:20:44 -0700" Date: Tue, 11 Jan 2005 14:59:55 -0500 Message-ID: <3615.1105473595@sss.pgh.pa.us> From: Tom Lane X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, hits=0.003 tagged_above=0 required=5 tests=AWL X-Spam-Level: X-Archive-Number: 200501/67 X-Sequence-Number: 15926 "Lee Wu" writes: > Here is postgres log from 20:17:42 to 20:25:35. > I did NOT removing anything except hostname and table names. Hmm. That shows the checkpoint process (8213) in a tight loop removing unwanted XLOG files. But apparently it is being starved for cycles --- look at the gaps between log entries. There's no problem with removing eight per second at 20:17:46, but after that there are periods of 10 to 20 seconds where it doesn't get to run at all. It really should have finished this phase in a few seconds, but instead it's still trying to do the removal seven minutes later when the COPY process fails. Apparently the COPY and VACUUM and whatever else is going on are able to create/use XLOG files faster than the checkpoint process can remove them. Which makes no sense. I think there is something fairly odd about either your kernel scheduler or your filesystem. What sort of platform is this anyway? regards, tom lane