From: "Ed L." <pgsql@bluepolka.net>
To: Tom Lane <tgl@sss.pgh.pa.us>
Subject: Re: query log corrupted-looking entries
Date: Fri, 1 Jun 2007 16:03:21 -0600
User-Agent: KMail/1.9.5
Cc: "George Pavlov" <gpavlov@mynewplace.com>,
 pgsql-general@postgresql.org
References: <29860.1161202644@sss.pgh.pa.us>
	<8C5B026B51B6854CBE88121DBF097A86CDD9A4@ehost010-33.exch010.intermedia.net>
	<24195.1180733773@sss.pgh.pa.us>
In-Reply-To: <24195.1180733773@sss.pgh.pa.us>
MIME-Version: 1.0
Content-Type: text/plain;
  charset="utf-8"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <200706011603.21322.pgsql@bluepolka.net>

On Friday 01 June 2007 3:36 pm, Tom Lane wrote:
> "George Pavlov" <gpavlov@mynewplace.com> writes:
> > On 5/29/2007 10:19 AM, Ed L. wrote:
> >> FWIW, I've also been seeing this sort of query log
> >> corruption for as long as I can remember, 7.1 through 8.2,
> >> HPUX (parisc, ia64), Linux on intel, amd...
>
> What *exactly* is the logging setup you guys use, and have you
> tried alternatives?
>
> As best I can tell, the variant with a downstream syslogger
> process ought to be OK for messages up to PIPEBUF bytes long
> on most standard Unixen.  Of course that's no help if you've
> got lots of very verbose queries, but the examples posted so
> far didn't seem to indicate that.

On older versions, we use stderr piped to Apache's rotatelogs.  
On newer versions with built-in logging, we use it.  We do have 
considerably verbose queries (sometimes a page or so).  No idea 
what PIPEBUF equals.

Ed