X-Original-To: pgsql-performance-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [200.46.204.2]) by svr1.postgresql.org (Postfix) with ESMTP id 110F9D1B510 for ; Tue, 21 Oct 2003 18:54:04 +0000 (GMT) Received: from svr1.postgresql.org ([200.46.204.71]) by localhost (neptune.hub.org [200.46.204.2]) (amavisd-new, port 10024) with ESMTP id 26521-10 for ; Tue, 21 Oct 2003 15:53:34 -0300 (ADT) Received: from davinci.ethosmedia.com (unknown [209.128.84.228]) by svr1.postgresql.org (Postfix) with ESMTP id 0BCC2D1B527 for ; Tue, 21 Oct 2003 15:53:29 -0300 (ADT) Received: from [66.219.92.2] (HELO temoku) by davinci.ethosmedia.com (CommuniGate Pro SMTP 4.0.2) with ESMTP id 3802455; Tue, 21 Oct 2003 11:54:00 -0700 Content-Type: text/plain; charset="iso-8859-1" From: Josh Berkus Reply-To: josh@agliodbs.com Organization: Aglio Database Solutions To: Andrew Sullivan , pgsql-performance@postgresql.org Subject: Re: Tuning for mid-size server Date: Tue, 21 Oct 2003 11:51:02 -0700 User-Agent: KMail/1.4.3 References: <2F2E24372F10744588A27DEECC85FE04B6769D@vt-pe2550-001.vantage.vantage.com> <200310211015.57569.josh@agliodbs.com> <20031021175016.GH11421@libertyrms.info> In-Reply-To: <20031021175016.GH11421@libertyrms.info> Cc: Bruce Momjian MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Message-Id: <200310211151.02751.josh@agliodbs.com> X-Virus-Scanned: by amavisd-new at postgresql.org X-Archive-Number: 200310/551 X-Sequence-Number: 4299 Andrew, > I think the "25%" rule of thumb is slightly stale: above some > threshold, it just falls apart, and lots of people now have machines > well within that threshold. Heck, I'll bet Bruce's 2-way machine is > within that threshold. Sure. But we had a few people on this list do tests (including me) and the= =20 anecdotal evidence was lower than 25%, substantially. The falloff is subt= le=20 until you hit 50% of RAM, like: % query throughput 1 ---- 5 --------- 10 ----------- 15 ---------- 20 ---------- 25 --------- 30 -------- 35 -------- 40 ------- ... so it's often not immediately apparent when you've set stuff a little t= oo=20 high. However, in the folks that tested, the ideal was never anywhere ne= ar=20 25%, usually more in the realm of 5-10%. I've been using 6% as my starting= =20 figure for the last year for a variety of servers with good results. Of course, if you have anecdotal evidence to the contrary, then the only wa= y=20 to work this would be to have OSDL help us sort it out. --=20 -Josh Berkus Aglio Database Solutions San Francisco