From owner-freebsd-performance@FreeBSD.ORG Wed Dec 1 17:13:59 2004 Return-Path: Delivered-To: freebsd-performance@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id F274916A4CE for ; Wed, 1 Dec 2004 17:13:59 +0000 (GMT) Received: from mail.trippynames.com (mail.trippynames.com [38.113.223.19]) by mx1.FreeBSD.org (Postfix) with ESMTP id DB9B743D49 for ; Wed, 1 Dec 2004 17:13:59 +0000 (GMT) (envelope-from sean@chittenden.org) Received: from localhost (localhost [127.0.0.1]) by mail.trippynames.com (Postfix) with ESMTP id 906E0A6CED; Wed, 1 Dec 2004 09:13:59 -0800 (PST) Received: from mail.trippynames.com ([127.0.0.1]) by localhost (rand.nxad.com [127.0.0.1]) (amavisd-new, port 10024) with LMTP id 97852-04; Wed, 1 Dec 2004 09:13:58 -0800 (PST) Received: from [192.168.102.101] (c-67-168-93-230.client.comcast.net [67.168.93.230]) by mail.trippynames.com (Postfix) with ESMTP id C568AA2D8D; Wed, 1 Dec 2004 09:13:57 -0800 (PST) In-Reply-To: <20041201105113.79348.qmail@web14102.mail.yahoo.com> References: <20041201105113.79348.qmail@web14102.mail.yahoo.com> Mime-Version: 1.0 (Apple Message framework v619) Content-Type: text/plain; charset=US-ASCII; format=flowed Message-Id: <6174636A-43BC-11D9-ACAF-000A95C705DC@chittenden.org> Content-Transfer-Encoding: 7bit From: Sean Chittenden Date: Wed, 1 Dec 2004 09:13:55 -0800 To: Claus Guttesen X-Mailer: Apple Mail (2.619) cc: freebsd-performance@freebsd.org Subject: Re: postgresql on FreeBSD 5.3 and high load X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 01 Dec 2004 17:14:00 -0000 > Our postgresql-server hits the wall when we get many > hits on our web. The load-average reaches 70. It > serves 11 webservers. Have you looked to see if you're using spin locks or not? Search for "amd64 x86_64 spin lock" in PostgreSQL's archives. I think spin locks were disabled for amd64. When you're at a load of 70, look in pg_catalog.pg_locks to see if you have any ungranted locks. Also, is your application primarily read? You may want to investigate using pgpool as a way of reducing the overhead for connection startup. Another option is to use memcached and remove some queries from the database all together. -sc -- Sean Chittenden