X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org
Received: from localhost (av.hub.org [200.46.204.144])
	by svr1.postgresql.org (Postfix) with ESMTP id 0A176D7CE2
	for <pgsql-performance-postgresql.org@localhost.postgresql.org>;
	Tue,  8 Nov 2005 15:10:55 -0400 (AST)
Received: from svr1.postgresql.org ([200.46.204.71])
	by localhost (av.hub.org [200.46.204.144]) (amavisd-new, port 10024)
	with ESMTP id 88771-03
	for <pgsql-performance-postgresql.org@localhost.postgresql.org>;
	Tue,  8 Nov 2005 19:10:52 +0000 (GMT)
X-Greylist: delayed 02:45:48.137469 by SQLgrey-
Received: from mail.Mi8.com (d01gw03.mi8.com [63.240.6.42])
	by svr1.postgresql.org (Postfix) with ESMTP id 0E14AD7B14
	for <pgsql-performance@postgresql.org>;
	Tue,  8 Nov 2005 15:10:52 -0400 (AST)
Received: from 172.16.1.112 by mail.Mi8.com with ESMTP (- Welcome to Mi8
	Corporation www.Mi8.com (D3)); Tue, 08 Nov 2005 14:10:43 -0500
X-Server-Uuid: E847189C-FC88-4913-9CD4-DE66914F83C0
Received: from MI8NYCMAIL06.Mi8.com ([172.16.1.175]) by
	D01SMTP02.Mi8.com with Microsoft SMTPSVC(6.0.3790.211); Tue, 8 Nov 2005
	14:09:09 -0500
Received: from 67.103.45.218 ([67.103.45.218]) by MI8NYCMAIL06.Mi8.com (
	[172.16.1.219]) via Exchange Front-End Server mi8owa.mi8.com (
	[172.16.1.104]) with Microsoft Exchange Server HTTP-DAV ; Tue, 8 Nov
	2005 14:09:08 -0500
User-Agent: Microsoft-Entourage/11.2.1.051004
Date: Tue, 08 Nov 2005 11:09:07 -0800
Subject: Re: Sort performance on large tables
From: "Luke Lonergan" <llonergan@greenplum.com>
To: "Stephan Szabo" <sszabo@megazone.bigpanda.com>,
	"Simon Riggs" <simon@2ndquadrant.com>,
	"Kurt Harriman" <kharriman@greenplum.com>
Cc: pgsql-performance@postgresql.org
Message-ID: <BF9637D3.1338F%llonergan@greenplum.com>
Thread-Topic: [PERFORM] Sort performance on large tables
Thread-Index: AcXki0Y6Ii/3K6rFQ2e9hfcWTSaFmAADJ39f
In-Reply-To: <20051108093500.N31541@megazone.bigpanda.com>
MIME-Version: 1.0
X-OriginalArrivalTime: 08 Nov 2005 19:09:09.0403 (UTC)
	FILETIME=[E5A6B6B0:01C5E497]
X-WSS-ID: 6F6E27B92BW4092657-09-01
Content-Type: multipart/alternative;
 boundary=B_3214292948_8404042
X-Virus-Scanned: by amavisd-new at hub.org
X-Spam-Status: No, score=1.408 required=5 tests=[AWL=0.154,
	HTML_MESSAGE=0.001, RCVD_NUMERIC_HELO=1.253]
X-Spam-Score: 1.408
X-Spam-Level: *
X-Archive-Number: 200511/143
X-Sequence-Number: 15400

> This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

--B_3214292948_8404042
Content-Type: text/plain;
 charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

Stephan,

On 11/8/05 9:38 AM, "Stephan Szabo" <sszabo@megazone.bigpanda.com> wrote:

>> >
>> > Just as we find with a similar comparison (with a "popular commercial,
>> > proprietary database" :-) Though some might suggest you increase
>> > work_mem or other tuning suggestions to speed sorting, none work.  In
>> > fact, we find that increasing work_mem actually slows sorting slightly=
.
>=20
> I wish you'd qualify your statements, because I can demonstrably show tha=
t
> I can make sorts go faster on my machine at least by increasing work_mem
> under some conditions.
>=20
Cool =AD can you provide your test case please?  I=B9ll ask our folks to do the
same, but as I recall we did some pretty thorough testing and found that it
doesn=B9t help.  Moreover, the conclusion was that the current algorithm isn=B9=
t
designed to use memory effectively.

Recognize also that we=B9re looking for a factor of 10 or more improvement
here =AD this is not a small increase that=B9s needed.

- Luke


--B_3214292948_8404042
Content-Type: text/html;
 charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

<HTML>
<HEAD>
<TITLE>Re: [PERFORM] Sort performance on large tables</TITLE>
</HEAD>
<BODY>
<FONT FACE=3D"Verdana, Helvetica, Arial"><SPAN STYLE=3D'font-size:14.0px'>Steph=
an,<BR>
<BR>
On 11/8/05 9:38 AM, &quot;Stephan Szabo&quot; &lt;sszabo@megazone.bigpanda.=
com&gt; wrote:<BR>
<BR>
</SPAN></FONT><BLOCKQUOTE><FONT FACE=3D"Verdana, Helvetica, Arial"><SPAN STYL=
E=3D'font-size:14.0px'>&gt;<BR>
&gt; Just as we find with a similar comparison (with a &quot;popular commer=
cial,<BR>
&gt; proprietary database&quot; :-) Though some might suggest you increase<=
BR>
&gt; work_mem or other tuning suggestions to speed sorting, none work. &nbs=
p;In<BR>
&gt; fact, we find that increasing work_mem actually slows sorting slightly=
.<BR>
<BR>
I wish you'd qualify your statements, because I can demonstrably show that<=
BR>
I can make sorts go faster on my machine at least by increasing work_mem<BR=
>
under some conditions.<BR>
<BR>
</SPAN></FONT></BLOCKQUOTE><FONT FACE=3D"Verdana, Helvetica, Arial"><SPAN STY=
LE=3D'font-size:14.0px'>Cool &#8211; can you provide your test case please? &n=
bsp;I&#8217;ll ask our folks to do the same, but as I recall we did some pre=
tty thorough testing and found that it doesn&#8217;t help. &nbsp;Moreover, t=
he conclusion was that the current algorithm isn&#8217;t designed to use mem=
ory effectively.<BR>
<BR>
Recognize also that we&#8217;re looking for a factor of 10 or more improvem=
ent here &#8211; this is not a small increase that&#8217;s needed.<BR>
<BR>
- Luke<BR>
</SPAN></FONT>
</BODY>
</HTML>


--B_3214292948_8404042--