Received: from candle.pha.pa.us (candle.navpoint.com [162.33.245.46]) by mail.postgresql.org (8.11.1/8.11.1) with ESMTP id eBIGNA694843 for ; Mon, 18 Dec 2000 11:23:11 -0500 (EST) (envelope-from pgman@candle.pha.pa.us) Received: (from pgman@localhost) by candle.pha.pa.us (8.9.0/8.9.0) id LAA26713; Mon, 18 Dec 2000 11:22:20 -0500 (EST) From: Bruce Momjian Message-Id: <200012181622.LAA26713@candle.pha.pa.us> Subject: Re: Best database structure for timely ordered values In-Reply-To: <18883.977151415@sss.pgh.pa.us> "from Tom Lane at Dec 18, 2000 09:56:55 am" To: Tom Lane Date: Mon, 18 Dec 2000 11:22:20 -0500 (EST) CC: Reiner Dassing , pgsql-sql@postgresql.org X-Mailer: ELM [version 2.4ME+ PL77 (25)] MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII X-Archive-Number: 200012/196 X-Sequence-Number: 530 > Reiner Dassing writes: > > The primary index must be the epoch. > > As there will be no deletion from this data I fear - due to the > > internal representation of B-trees - the performance will degrade very > > soon. > > Nonsense. btree should work just fine for that. Use a timestamp > column for the primary key, and away you go. > > (Actually, time alone doesn't seem like it'd be necessarily unique, > so maybe you don't want to call it a primary key. But certainly > you can make a non-unique index on that column.) I assume the user is concerned about non-balanced btrees. Ours are auto-balancing. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania 19026