Date: Thu, 14 Apr 2005 15:04:47 -0400
From: Alvaro Herrera <alvherre@dcc.uchile.cl>
To: Andrew Sullivan <ajs@crankycanuck.ca>
Cc: pgsql-sql@postgresql.org
Subject: Re: row-attribute in EXPLAIN-output doesn't match count(*)
Message-ID: <20050414190447.GE28198@dcc.uchile.cl>
References: <200504141838.14705.andreak@officenet.no>
	<200504141921.38476.andreak@officenet.no>
	<20050414173358.GB9027@phlogiston.dyndns.org>
	<200504141954.20520.andreak@officenet.no>
	<20050414185623.GC9027@phlogiston.dyndns.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <20050414185623.GC9027@phlogiston.dyndns.org>
User-Agent: Mutt/1.5.8i

On Thu, Apr 14, 2005 at 02:56:23PM -0400, Andrew Sullivan wrote:
> On Thu, Apr 14, 2005 at 07:54:19PM +0200, Andreas Joseph Krogh wrote:
> > Any hints on how I can find out what's keeping the connection idle in a 
> > transaction? I realize now that I should probably ask that question on the 
> > pgsql-jdbc-list:-)
> 
> Nope.  That's a problem with your pool software.  It's no doubt
> issuing "BEGIN;" as soon as it connects.  It's BAD (broken as
> designed) in that case.  You need to convince it not to do that, or
> else you need to go around and kill -2 such connections from time to
> time.

It is a Postgres limitation as well.  We _could_ make the server "really
start the transaction" at the point the first query is issued instead of
when the BEGIN is issued.  In fact this problem would go away if we did
that.

-- 
Alvaro Herrera (<alvherre[@]dcc.uchile.cl>)
Voy a acabar con todos los humanos / con los humanos yo acabaré
voy a acabar con todos / con todos los humanos acabaré (Bender)