Date: Thu, 28 Aug 2008 08:36:46 -0400
From: Andrew Sullivan <ajs@commandprompt.com>
To: pgsql-performance@postgresql.org
Subject: Re: select on 22 GB table causes "An I/O error occured
	while sending to the backend." exception
Message-ID: <20080828123645.GB23588@commandprompt.com>
Mail-Followup-To: Andrew Sullivan <ajs@commandprompt.com>,
	pgsql-performance@postgresql.org
References: <BAY124-W57D3654AF1159CAF8EAC19F5660@phx.gbl>
	<82bpze3ou8.fsf@mid.bfk.de>
	<alpine.DEB.1.10.0808271436210.907@asgard.lang.hm>
	<20080827221338.GP18946@commandprompt.com>
	<alpine.DEB.1.10.0808271516300.907@asgard.lang.hm>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <alpine.DEB.1.10.0808271516300.907@asgard.lang.hm>
User-Agent: Mutt/1.5.17 (2007-11-01)

On Wed, Aug 27, 2008 at 03:22:09PM -0700, david@lang.hm wrote:
>
> I disagree with you. I think goof Postgres operation is so highly dependant 
> on caching as much data as possible that disabling overcommit (and throwing 
> away a lot of memory that could be used for cache) is a solution that's as 
> bad or worse than the problem it's trying to solve.

Ok, but the danger is that the OOM killer kills your postmaster.  To
me, this is a cure way worse than the disease it's trying to treat.
YMMD &c. &c.

A

-- 
Andrew Sullivan
ajs@commandprompt.com
+1 503 667 4564 x104
http://www.commandprompt.com/