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 D460CD1C9F9 for ; Fri, 1 Aug 2003 21:19:54 +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 71975-06 for ; Fri, 1 Aug 2003 18:18:48 -0300 (ADT) Received: from davinci.ethosmedia.com (209-128-84-228.bayarea.net [209.128.84.228]) by svr1.postgresql.org (Postfix) with ESMTP id D3916D1C9F5 for ; Fri, 1 Aug 2003 18:19:52 -0300 (ADT) Received: from [66.219.92.2] (HELO temoku) by davinci.ethosmedia.com (CommuniGate Pro SMTP 4.0.2) with ESMTP id 3368865; Fri, 01 Aug 2003 14:20:15 -0700 Content-Type: text/plain; charset="iso-8859-1" From: Josh Berkus Reply-To: josh@agliodbs.com Organization: Aglio Database Solutions To: Francisco J Reyes Subject: Re: How number of columns affects performance Date: Fri, 1 Aug 2003 14:18:51 -0700 User-Agent: KMail/1.4.3 Cc: Ron Johnson , PostgreSQL performance References: <20030801115257.I92555@zoraida.natserv.net> <200308011044.03493.josh@agliodbs.com> <20030801163346.V49062@zoraida.natserv.net> In-Reply-To: <20030801163346.V49062@zoraida.natserv.net> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Message-Id: <200308011418.51572.josh@agliodbs.com> X-Virus-Scanned: by amavisd-new at postgresql.org X-Archive-Number: 200308/18 X-Sequence-Number: 2835 Francisco, > I think I understand your point, however it would be very laborious after > you do all development to find out you need to de-normalize. Not terribly. Views and Rules are good for this. > On your experience at which point it would actually help to do this > de-normalization in PostgreSQL? I know there are numerous factors ,but any > feedback based on previous experiences would help. My experience? If you're running on good hardware, it's completely=20 unnecessary to vertically partition the table. The only thing I'd do woul= d=20 be to look for columns which are frequently NULL and can be grouped togethe= r,=20 and spin those off into a sub-table. That is, if you have 4 columns which= =20 are generally either all null or all filled, and are all null for 70% of=20 records then those 4 could make a nice child table. --=20 -Josh Berkus ______AGLIO DATABASE SOLUTIONS___________________________ Josh Berkus Complete information technology josh@agliodbs.com and data management solutions (415) 565-7293 for law firms, small businesses fax 621-2533 and non-profit organizations. San Francisco