X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [64.117.224.130]) by svr1.postgresql.org (Postfix) with ESMTP id 3DA4AD1C4D0 for ; Thu, 31 Jul 2003 14:43:24 +0000 (GMT) Received: from svr1.postgresql.org ([64.117.224.193]) by localhost (neptune.hub.org [64.117.224.130]) (amavisd-new, port 10024) with ESMTP id 54792-01 for ; Thu, 31 Jul 2003 11:42:23 -0300 (ADT) Received: from snowwhite.lulu.com (unknown [66.193.5.50]) by svr1.postgresql.org (Postfix) with ESMTP id 0E83ED1C9D8 for ; Thu, 31 Jul 2003 11:43:21 -0300 (ADT) Received: from lulu.com (foo.rdu.lulu.com [10.0.0.123]) by snowwhite.lulu.com (8.11.6/8.11.6) with ESMTP id h6VEh0V15082 for ; Thu, 31 Jul 2003 10:43:15 -0400 Message-ID: <3F292AF4.1000104@lulu.com> Date: Thu, 31 Jul 2003 10:43:00 -0400 From: cafweb User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030701 X-Accept-Language: en-us, en MIME-Version: 1.0 To: pgsql-performance@postgresql.org Subject: Re: postgresql.conf References: In-Reply-To: Content-Type: multipart/alternative; boundary="------------000409090906070601000706" X-Virus-Scanned: by amavisd-new at postgresql.org X-Archive-Number: 200307/499 X-Sequence-Number: 2792 This is a multi-part message in MIME format. --------------000409090906070601000706 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit >So, maybe just a note on which parameters to increase if you have more >RAM/CPU/I/O bandwidth in the big server example? > > Yes, that would be great. Actually I prefer rules of thumb and examples for each extreme. If possible a little note WHY the parameter should be tweaked, and what effects it will have. For example, more note's like Josh's below would be a big help... >Yes. Take sort_mem for example: >OLTP_SM 1024 >OLTP_LM 2048 >OLAP_SM 4096 >OLAP_LM 16384 >WWW_SM 512 >WWW_LM 1024 >Workstation 1024 > >The basic idea is: >More RAM => more sort_mem >More concurrent queries => less sort_mem >Larger data sets => more sort_mem >Lots of grouped aggregates => more sort_mem > > > -- [ Christian Fowler [ cafweb@lulu.com [ http://www.lulu.com --------------000409090906070601000706 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit
So, maybe just a note on which parameters to increase if you have more 
RAM/CPU/I/O bandwidth in the big server example?
  
Yes, that would be great. Actually I prefer rules of thumb and examples for each extreme. If possible a little note WHY the parameter should be tweaked, and what effects it will have. For example, more note's like Josh's below would be a big help...

Yes. Take sort_mem for example:
OLTP_SM	1024
OLTP_LM	2048
OLAP_SM	4096
OLAP_LM	16384
WWW_SM	512
WWW_LM	1024
Workstation	1024

The basic idea is:
More RAM => more sort_mem
More concurrent queries => less sort_mem
Larger data sets => more sort_mem
Lots of grouped aggregates => more sort_mem

  

-- 
[ Christian Fowler
[ cafweb@lulu.com
[ http://www.lulu.com
--------------000409090906070601000706--