Date: Thu, 7 Jun 2007 11:45:50 -0400
From: Andrew Sullivan <ajs@crankycanuck.ca>
To: pgsql-performance@postgresql.org
Subject: Re: VERY slow queries at random
Message-ID: <20070607154550.GF14087@phlogiston.dyndns.org>
References: <46670916.9090505@gmail.com>
	<20070606195649.GG11545@phlogiston.dyndns.org>
	<466814B7.6000606@gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <466814B7.6000606@gmail.com>
User-Agent: Mutt/1.5.9i

On Thu, Jun 07, 2007 at 04:22:47PM +0200, Gunther Mayer wrote:
> There are a whole bunch of update queries that fire all the time but 
> afaik none of them ever lock the entire table. To the best of my 
> knowledge UPDATE ... WHERE ... only locks those rows that it actually 
> operates on, in my case this is always a single row.

Well that shouldn't be biting you, then (you're not in SERIALIZABLE
mode, right?).  The other obvious bit would be checkpoint storms. 
What's your bgwriter config like?

> Question is, how do I find out about locks at the time when I only get 
> told about the slow query *after* it has completed and postgres has told 
> me so by logging a slow query entry in my logs?

You can't :(

A

-- 
Andrew Sullivan  | ajs@crankycanuck.ca
This work was visionary and imaginative, and goes to show that visionary
and imaginative work need not end up well. 
		--Dennis Ritchie