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 2B46BD76BA
	for <pgsql-performance-postgresql.org@localhost.postgresql.org>;
	Tue, 13 Sep 2005 00:53:58 -0300 (ADT)
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 22316-05
	for <pgsql-performance-postgresql.org@localhost.postgresql.org>;
	Tue, 13 Sep 2005 03:53:56 +0000 (GMT)
Received: from wproxy.gmail.com (wproxy.gmail.com [64.233.184.206])
	by svr1.postgresql.org (Postfix) with ESMTP id 5C2A4D6E69
	for <pgsql-performance@postgresql.org>;
	Tue, 13 Sep 2005 00:53:55 -0300 (ADT)
Received: by wproxy.gmail.com with SMTP id 36so2357630wra
	for <pgsql-performance@postgresql.org>;
	Mon, 12 Sep 2005 20:53:55 -0700 (PDT)
DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com;
	h=received:message-id:date:from:reply-to:to:subject:in-reply-to:mime-version:content-type:references;
	b=FnTFga/ej/E7rLKPEIWK1mwq5AHBfW5ctFrA8emT/fGEBSw3IzbFebTdtgyqa3uHgQsbbsK/YUC5Nf/fucj5tU36SUjEh5GQFWKf1wnUVvtKHwTIQe6aOQxb+JrdBJVqwF2oH0rkePmwmsoBzTs7+t1748W4JKbQye7lMGmNP3A=
Received: by 10.54.21.44 with SMTP id 44mr190289wru;
	Mon, 12 Sep 2005 20:53:55 -0700 (PDT)
Received: by 10.54.158.11 with HTTP; Mon, 12 Sep 2005 20:53:55 -0700 (PDT)
Message-ID: <84621a605091220533aa8fa12@mail.gmail.com>
Date: Mon, 12 Sep 2005 22:53:55 -0500
From: Brandon Black <blblack@gmail.com>
Reply-To: blblack@gmail.com
To: petrilli@gmail.com, pgsql-performance@postgresql.org
Subject: Re: Performance considerations for very heavy INSERT traffic
In-Reply-To: <59d991c40509122039773cac17@mail.gmail.com>
Mime-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_Part_12393_22884737.1126583635048"
References: <84621a6050912140464c380c3@mail.gmail.com>
	<59d991c40509122039773cac17@mail.gmail.com>
X-Virus-Scanned: by amavisd-new at hub.org
X-Spam-Status: No, hits=0.139 required=5 tests=[AWL=0.058, HTML_30_40=0.056,
	HTML_MESSAGE=0.001, RCVD_BY_IP=0.024]
X-Spam-Level: 
X-Archive-Number: 200509/155
X-Sequence-Number: 14463

------=_Part_12393_22884737.1126583635048
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

On 9/12/05, Christopher Petrilli <petrilli@gmail.com> wrote:
>=20
>=20
> 2) Tune ext3. The default configuration wrecks high-write situations.
> Look into data=3Dwriteback for mounting, turning off atime (I hope
> you've done this already) updates, and also modifying the scheduler to
> the elevator model. This is poorly documented in Linux (like just
> about everything), but it's crtical.


I'm using noatime and data=3Dwriteback already. I changed my scheduler from=
=20
the default anticipatory to deadline and saw an improvement, but I haven't=
=20
yet delved into playing with specific elevator tunable values per-device.

3) Use 8.1 and strongly look at Bizgres. The data partitioning is critical.


I've just started down the path of getting 8.1 running with a larger block=
=20
size, and tommorow I'm going to look at Bizgres's partitioning as opposed t=
o=20
some manual schemes. Will the current Bizgres have a lot of the performance=
=20
enhancements of 8.1 already (or will 8.1 or 8.2 eventually get Bizgres's=20
partitioning?)?

-- Brandon

------=_Part_12393_22884737.1126583635048
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

<br><br><div><span class=3D"gmail_quote">On 9/12/05, <b class=3D"gmail_send=
ername">Christopher Petrilli</b> &lt;<a href=3D"mailto:petrilli@gmail.com">=
petrilli@gmail.com</a>&gt; wrote:</span><blockquote class=3D"gmail_quote" s=
tyle=3D"border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8e=
x; padding-left: 1ex;">
<br>2) Tune ext3.&nbsp;&nbsp;The default configuration wrecks high-write si=
tuations.<br> Look into data=3Dwriteback for mounting, turning off atime (I=
 hope<br>you've done this already) updates, and also modifying the schedule=
r to<br>
the elevator model.&nbsp;&nbsp;This is poorly documented in Linux (like jus=
t<br>about everything), but it's crtical.</blockquote><div><br>
I'm using noatime and data=3Dwriteback already.&nbsp; I changed my
scheduler from the default anticipatory to deadline and saw an
improvement, but I haven't yet delved into playing with specific
elevator tunable values per-device.<br>
</div><br><blockquote class=3D"gmail_quote" style=3D"border-left: 1px solid=
 rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">3) Use =
8.1 and strongly look at Bizgres. The data partitioning is critical.</block=
quote>
<div><br>
I've just started down the path of getting 8.1 running with a larger
block size, and tommorow I'm going to look at Bizgres's partitioning as
opposed to some manual schemes.&nbsp; Will the current Bizgres have a
lot of the performance enhancements of 8.1 already (or will 8.1 or 8.2
eventually get Bizgres's partitioning?)?<br>
</div><br>-- Brandon<br>
</div>

------=_Part_12393_22884737.1126583635048--