Message-ID: <4FD5FD5C.5040203@optionshouse.com>
Date: Mon, 11 Jun 2012 09:14:52 -0500
From: Shaun Thomas <sthomas@optionshouse.com>
Reply-To: <sthomas@optionshouse.com>
Organization: Peak6
User-Agent: Mozilla/5.0 (X11; Linux i686;
	rv:12.0) Gecko/20120430 Thunderbird/12.0.1
MIME-Version: 1.0
To: Mark Thornton <mthornton@optrak.com>
CC: <pgsql-performance@postgresql.org>
Subject: Re: Performance of CLUSTER
References: <4FD458CA.3070902@optrak.com> <4FD5F5DC.3080507@optionshouse.com>
	<4FD5F6B5.2090505@optrak.com> <4FD5F837.5080006@optionshouse.com>
	<4FD5FA6F.3080903@optrak.com>
In-Reply-To: <4FD5FA6F.3080903@optrak.com>
Content-Type: text/plain; charset="ISO-8859-1"; format=flowed
content-transfer-encoding: 7bit

On 06/11/2012 09:02 AM, Mark Thornton wrote:

> I didn't think the process was using even the 500m it ought to have
> had available, whereas creating an index did appear to use that much.
> Note though that I didn't stay up all night watching it!

You'd be surprised. If you look in your base/pgsql_tmp directory during 
a cluster of that table (make a copy of it if you don't want to 
interfere with a running system) you should see that directory fill with 
temporary structures, mostly during the index rebuild portions.

It also wouldn't hurt to bootstrap system cache with the contents of 
that table. Do an explain analyze on SELECT * with no where clause and 
all of that table should be in memory.

Oh, actually that reminds me... does your 10GB table fit into memory? If 
not, that might explain it right there.

-- 
Shaun Thomas
OptionsHouse | 141 W. Jackson Blvd. | Suite 500 | Chicago IL, 60604
312-444-8534
sthomas@optionshouse.com

______________________________________________

See http://www.peak6.com/email_disclaimer/ for terms and conditions related to this email