Received: from sss.pgh.pa.us (unknown [192.204.191.242]) by postgresql.org (Postfix) with ESMTP id 0E8EF4759FA for ; Fri, 5 Apr 2002 18:50:18 -0500 (EST) Received: from sss2.sss.pgh.pa.us (tgl@localhost [127.0.0.1]) by sss.pgh.pa.us (8.11.4/8.11.4) with ESMTP id g35NoHF11104; Fri, 5 Apr 2002 18:50:17 -0500 (EST) To: d_nardini@btconnect.com Cc: pgsql-sql@postgresql.org Subject: Re: intersect performance (PG 7.1.3 vs 7.2) In-reply-to: <0CBD3E12-48E6-11D6-8010-0030654E696C@btconnect.com> References: <0CBD3E12-48E6-11D6-8010-0030654E696C@btconnect.com> Comments: In-reply-to d_nardini@btconnect.com message dated "Fri, 05 Apr 2002 22:39:54 +0000" Date: Fri, 05 Apr 2002 18:50:16 -0500 Message-ID: <11101.1018050616@sss.pgh.pa.us> From: Tom Lane X-Archive-Number: 200204/76 X-Sequence-Number: 7341 d_nardini@btconnect.com writes: > I'm running the following statement on a (development) Mac OS X (PG 7.2) > box and on a (production) RedHat 7.0 (PG 7.1.3) box. On the OS X it's > performing very well ... on RH it's VERY SLOW, with CPU utilization > hitting 90%+ (on OS X it hardly exceeds 2%). Hmm. Nested intersects are actually broken in 7.1.*, cf. http://archives.postgresql.org/pgsql-bugs/2001-08/msg00064.php Do you get the same results from both boxes? > BTW - replacing 'intersect' > with 'union' in the SQL behaves the same. In that case it might possibly be a different problem. Do you get the same EXPLAIN plan on both boxes? regards, tom lane