X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [200.46.204.144]) by svr1.postgresql.org (Postfix) with ESMTP id C1A0D542F3 for ; Tue, 10 May 2005 14:31:32 -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 23340-05 for ; Tue, 10 May 2005 17:31:25 +0000 (GMT) Received: from boutiquenumerique.com (unknown [82.67.9.10]) by svr1.postgresql.org (Postfix) with ESMTP id 79F23540DA for ; Tue, 10 May 2005 14:31:21 -0300 (ADT) Received: (qmail 10820 invoked from network); 10 May 2005 19:30:22 +0200 Received: from unknown (HELO localhost) (boutiquenumerique-lists@192.168.0.4) by boutiquenumerique.com with SMTP; 10 May 2005 19:30:22 +0200 To: "John A Meinel" , "Adam Haberlach" Cc: pgsql-performance@postgresql.org Subject: Re: Partitioning / Clustering References: <20050510150226.00E453EFF@flute.newsnipple.com> <4280D76E.3090209@arbash-meinel.com> Message-ID: Date: Tue, 10 May 2005 19:29:59 +0200 From: PFC Organization: =?iso-8859-15?Q?La_Boutique_Num=E9rique?= Content-Type: text/plain; format=flowed; delsp=yes; charset=iso-8859-15 MIME-Version: 1.0 Content-Transfer-Encoding: 8bit In-Reply-To: <4280D76E.3090209@arbash-meinel.com> User-Agent: Opera M2(BETA2)/8.0 (Linux, build 987) X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, hits=0.002 tagged_above=0 required=5 tests=AWL X-Spam-Level: X-Archive-Number: 200505/152 X-Sequence-Number: 12316 > SELECT row1, row2 FROM table1_on_machine_a NATURAL JOIN > table2_on_machine_b > WHERE restrict_table_1 AND restrict_table_2 > AND restrict_1_based_on_2; I don't think that's ever going to be efficient... What would be efficient would be, for instance, a Join of a part of a table against another part of another table which both happen to be on the same machine, because the partitioning was done with this in mind (ie. for instance partitioning on client_id and keeping the information for each client on the same machine). You could build your smart pool daemon in pl/pgsql and use dblink ! At least you have the query parser built-in. I wonder how Oracle does it ;)