Date: Wed, 8 Aug 2012 13:11:56 -0400
From: Wayne Cuddy <lists-pgsql@useunix.net>
To: PostgreSQL <pgsql-sql@postgresql.org>
Subject: Re: fsync debug messages in pgsql logs
Message-ID: <20120808171156.GG16371@slacker.ja10629.home>
Mail-Followup-To: PostgreSQL <pgsql-sql@postgresql.org>
References: <20120808154416.GE16371@slacker.ja10629.home>
	<12758.1344443002@sss.pgh.pa.us>
	<20120808165304.GF16371@slacker.ja10629.home>
	<13444.1344445021@sss.pgh.pa.us>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <13444.1344445021@sss.pgh.pa.us>
User-Agent: Mutt/1.4.2.3i

Ok, I'll keep an eye on it but I'm not so worried now.

Thanks Tom.


On Wed, Aug 08, 2012 at 12:57:01PM -0400, Tom Lane wrote:
> Wayne Cuddy <lists-pgsql@useunix.net> writes:
> > On Wed, Aug 08, 2012 at 12:23:22PM -0400, Tom Lane wrote:
> >> If it only complains once per file name, this is expected behavior when
> >> somebody drops a table just before the checkpoint mechanism tries to
> >> fsync it.  (If the failure were to repeat, then it might be something
> >> more interesting.)  It does seem a bit odd that only fsm files are being
> >> complained of, though.
> >> 
> >> What PG version is that exactly?
> 
> > It's 9.0.4. I'm frequently truncating tables but not dropping them.
> 
> Under the hood, TRUNCATE creates a new empty table and then drops the
> old one after commit.  So that sounds consistent.  It's still a bit odd
> that all the messages are about FSM files, but maybe that has something
> to do with sequence-of-operations in the DROP.  Anyway, nothing to see
> here AFAICT.
> 
> 			regards, tom lane
> 
> -- 
> Sent via pgsql-sql mailing list (pgsql-sql@postgresql.org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-sql