X-Original-To: pgsql-hackers-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [200.46.204.144]) by svr1.postgresql.org (Postfix) with ESMTP id 19E41539D5 for ; Tue, 26 Apr 2005 20:13:11 -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 47602-01 for ; Tue, 26 Apr 2005 23:13:06 +0000 (GMT) Received: from server2.sitesell.com (server2.sitesell.com [216.95.221.3]) by svr1.postgresql.org (Postfix) with SMTP id 465165377C for ; Tue, 26 Apr 2005 20:13:03 -0300 (ADT) Received: (qmail 6217 invoked by uid 223); 26 Apr 2005 23:13:03 -0000 Received: from 216.154.15.113 by server2.sitesell.com (envelope-from , uid 101) with qmail-scanner-1.25 (clamdscan: 0.81/685. spamassassin: 3.0.1. Clear:RC:1(216.154.15.113):. Processed in 0.027557 secs); 26 Apr 2005 23:13:03 -0000 Received: from 216-154-15-113.dsl.look.ca (216-154-15-113.dsl.look.ca [216.154.15.113]) by server2.sitesell.com ([216.95.221.3]) with ESMTP via TCP; 26 Apr 2005 23:13:03 -0000 Subject: Re: [PERFORM] Bad n_distinct estimation; hacks suggested? From: Rod Taylor To: Greg Stark Cc: Gurmeet Manku , pgsql-hackers@postgresql.org In-Reply-To: <87u0lt5efo.fsf@stark.xeocode.com> References: <87u0lt5efo.fsf@stark.xeocode.com> Content-Type: text/plain Date: Tue, 26 Apr 2005 19:10:11 -0400 Message-Id: <1114557011.77587.107.camel@home> Mime-Version: 1.0 X-Mailer: Evolution 2.2.2 FreeBSD GNOME Team Port Content-Transfer-Encoding: 7bit X-Virus-Scanned: by amavisd-new at hub.org X-Spam-Status: No, hits=0 tagged_above=0 required=5 tests= X-Spam-Level: X-Archive-Number: 200504/894 X-Sequence-Number: 66763 On Tue, 2005-04-26 at 19:03 -0400, Greg Stark wrote: > This one looks *really* good. > > http://www.aladdin.cs.cmu.edu/papers/pdfs/y2001/dist_sampl.pdf > > It does require a single full table scan Ack.. Not by default please. I have a few large append-only tables (vacuum isn't necessary) which do need stats rebuilt periodically. Lets just say that we've been working hard to upgrade to 8.0 primarily because pg_dump was taking over 18 hours to make a backup. -- Rod Taylor