Message-Id: <1166808625.17094.281711491@webmail.messagingengine.com>
From: "Jeremy Haile" <jhaile@fastmail.fm>
To: "Simon Riggs" <simon@2ndquadrant.com>
Cc: pgsql-performance@postgresql.org
Content-Disposition: inline
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset="ISO-8859-1"
MIME-Version: 1.0
References: <1166803606.2828.281701761@webmail.messagingengine.com>
	<1166806378.10592.281708161@webmail.messagingengine.com>
	<1166806963.3783.203.camel@silverbirch.site>
Subject: Re: URGENT: Out of disk space pg_xlog
In-Reply-To: <1166806963.3783.203.camel@silverbirch.site>
Date: Fri, 22 Dec 2006 12:30:25 -0500

The archive_status directory is empty.  I've never seen any files in
there and I've never set archive_command.

Well, the problem has since resolved, but here is what is in the
directory now.  Previously there were hundreds of files, but these
disappeared after Postgres performed the automatic recovery.

12/22/2006  11:16 AM        16,777,216 0000000100000030000000D2
12/22/2006  11:17 AM        16,777,216 0000000100000030000000D3
12/22/2006  11:17 AM        16,777,216 0000000100000030000000D4
12/22/2006  11:17 AM        16,777,216 0000000100000030000000D5
12/22/2006  11:18 AM        16,777,216 0000000100000030000000D6
12/22/2006  11:19 AM        16,777,216 0000000100000030000000D7
12/22/2006  11:19 AM        16,777,216 0000000100000030000000D8
12/22/2006  11:19 AM        16,777,216 0000000100000030000000D9
12/22/2006  11:19 AM        16,777,216 0000000100000030000000DA
12/22/2006  11:21 AM        16,777,216 0000000100000030000000DB
12/22/2006  10:07 AM        16,777,216 0000000100000030000000DC
12/22/2006  10:07 AM        16,777,216 0000000100000030000000DD
12/22/2006  10:07 AM        16,777,216 0000000100000030000000DE
12/22/2006  10:33 AM        16,777,216 0000000100000030000000DF
12/22/2006  10:08 AM        16,777,216 0000000100000030000000E0
12/22/2006  10:32 AM        16,777,216 0000000100000030000000E1
12/22/2006  10:08 AM        16,777,216 0000000100000030000000E2
12/22/2006  10:08 AM        16,777,216 0000000100000030000000E3
12/22/2006  10:17 AM        16,777,216 0000000100000030000000E4
12/22/2006  10:11 AM        16,777,216 0000000100000030000000E5
12/22/2006  11:10 AM        16,777,216 0000000100000030000000E6
12/22/2006  11:11 AM        16,777,216 0000000100000030000000E7
12/22/2006  11:15 AM        16,777,216 0000000100000030000000E8
12/22/2006  11:15 AM        16,777,216 0000000100000030000000E9
12/22/2006  11:15 AM        16,777,216 0000000100000030000000EA
12/22/2006  11:16 AM        16,777,216 0000000100000030000000EB
12/22/2006  11:16 AM        16,777,216 0000000100000030000000EC
12/22/2006  11:16 AM        16,777,216 0000000100000030000000ED
12/18/2006  08:52 PM    <DIR>          archive_status
              28 File(s)    469,762,048 bytes
               3 Dir(s)  10,206,756,864 bytes free

On Fri, 22 Dec 2006 17:02:43 +0000, "Simon Riggs"
<simon@2ndquadrant.com> said:
> On Fri, 2006-12-22 at 11:52 -0500, Jeremy Haile wrote:
> 
> > I would still appreciate ... explaining how to estimate size for a
> > pg_xlog partition.  It seems like it can vary considerably depending on
> > how intensive your current transactions are.  Is there a way to
> > determine a maximum?
> 
> There should be at most 2*checkpoint_segments+1 files in pg_xlog, which
> are 16MB each. So you shouldn't be having a problem.
> 
> If there are more than this, it could be because you have
> currently/previously had archive_command set and the archive command
> failed to execute correctly, or the database was shutdown/crashed prior
> to the archive commands being executed.
> 
> IIRC there was a bug that allowed this to happen, but that was some time
> ago.
> 
> Perhaps you could show us the dir listing, so we can check that there is
> not a new problem emerging? Can you also show us the contents of the
> pg_xlog/archive_status directory? Thanks.
> 
> -- 
>   Simon Riggs             
>   EnterpriseDB   http://www.enterprisedb.com
> 
>