X-Original-To: pgsql-sql-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [200.46.204.144]) by svr1.postgresql.org (Postfix) with ESMTP id 28E3352881 for ; Wed, 8 Jun 2005 07:30:25 -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 77602-10 for ; Wed, 8 Jun 2005 10:30:16 +0000 (GMT) Received: from mx-00.sil.at (mx-00.sil.at [62.116.68.196]) by svr1.postgresql.org (Postfix) with ESMTP id 7CF095283C for ; Wed, 8 Jun 2005 07:30:12 -0300 (ADT) Received: (qmail-ldap/ctrl 46563 invoked from network); 8 Jun 2005 10:30:10 -0000 Received: from unknown (HELO io.sonorys.at) ([213.235.243.42]) (envelope-sender ) by mx-00.sil.at (qmail-ldap-1.03) with SMTP for ; 8 Jun 2005 10:30:10 -0000 Received: from neptun.sonorys.at (neptun.sonorys.at [172.24.1.17]) by io.sonorys.at (8.11.7+Sun/8.11.7) with ESMTP id j58AUBG04548 for ; Wed, 8 Jun 2005 12:30:11 +0200 (MEST) Received: by neptun.sonorys.at with Internet Mail Service (5.5.2653.19) id ; Wed, 8 Jun 2005 12:30:11 +0200 Message-ID: From: =?ISO-8859-1?Q?K=D6PFERL_Robert?= To: Cc: pgsql-sql@postgresql.org Subject: Experience with splitting a Table transparently Date: Wed, 8 Jun 2005 12:30:10 +0200 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain; charset="ISO-8859-1" X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, hits=0.015 tagged_above=0 required=5 tests=AWL X-Spam-Level: X-Archive-Number: 200506/109 X-Sequence-Number: 21710 Hi, I'm looking for people who have got experience at splitting a table with heavy records into two relations. In my case there exists one table with the mentioned heavy records. These are processed by a statemachine. Thus a bunch of columns gets changed several times in the livetime of one record. The text and blob colmns remain unchanged. However due to Postgres# MVCC one changed record gets duplicated, even if just one bit is changed. This makes us IO-performance problems. I intend now to split the table in transport data (which stays constant) and status data (which is changed several times). I have then several options: Change all stored procedures to mathc the new relations and keep the interface equal. Or create a view with a bunch of RULES to show the procedures the 'original table'. What is best? Who has done such bevore? Will this help us in IO-terms? thanks