X-Original-To: pgsql-performance@postgresql.org Received: from spampd.localdomain (postgresql.org [64.49.215.8]) by postgresql.org (Postfix) with ESMTP id 0DAAB475E77 for ; Fri, 25 Apr 2003 12:28:35 -0400 (EDT) Received: from fuji.krosing.net (sein.itera.ee [194.126.109.126]) by postgresql.org (Postfix) with ESMTP id 0C922474E42 for ; Fri, 25 Apr 2003 12:28:34 -0400 (EDT) Received: from fuji.krosing.net (lo [127.0.0.1]) by fuji.krosing.net (8.12.8/8.12.8) with ESMTP id h3PGSIJ0002719; Fri, 25 Apr 2003 19:28:18 +0300 Received: (from hannu@localhost) by fuji.krosing.net (8.12.8/8.12.8/Submit) id h3PGS6Ss002717; Fri, 25 Apr 2003 19:28:06 +0300 X-Authentication-Warning: fuji.krosing.net: hannu set sender to hannu@tm.ee using -f Subject: Re: More tablescanning fun From: Hannu Krosing To: jim@nasby.net Cc: Tom Lane , pgsql-performance@postgresql.org In-Reply-To: <20030424235924.B66185@flake.decibel.org> References: <20030424183817.A66185@flake.decibel.org> <3660.1051228710@sss.pgh.pa.us> <20030424235924.B66185@flake.decibel.org> Content-Type: text/plain Content-Transfer-Encoding: 7bit Organization: Message-Id: <1051288086.1765.95.camel@fuji.krosing.net> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.2.4 Date: 25 Apr 2003 19:28:06 +0300 X-Spam-Status: No, hits=-32.9 required=5.0 tests=BAYES_10,IN_REP_TO,QUOTED_EMAIL_TEXT,REFERENCES, REPLY_WITH_QUOTES,USER_AGENT_XIMIAN,X_AUTH_WARNING autolearn=ham version=2.50 X-Spam-Level: X-Spam-Checker-Version: SpamAssassin 2.50 (1.173-2003-02-20-exp) X-Archive-Number: 200304/210 X-Sequence-Number: 1716 Jim C. Nasby kirjutas R, 25.04.2003 kell 07:59: > > > Also, is there a TODO to impliment > > > real clustered indexes? > > > > No. It's not apparent to me how you could do that without abandoning > > MVCC, which we're not likely to do. > > Hmm... does MVCC mandate inserts go at the end? I have been pondering if keeping pages half-empty (or even 70% empty) could solve both clustering problems and longish updates for much data. If we could place the copy in the same page than original, most big updates would be possible by one sweep of disk heads and also clustering order would be easier to keep if pages were kept intentionally half empty. So "VACUUM FULL 65% EMPTY;" could make sense ? ------------- Hannu