From owner-freebsd-database@FreeBSD.ORG Mon Feb 6 16:59:12 2006 Return-Path: X-Original-To: freebsd-database@freebsd.org Delivered-To: freebsd-database@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id D92F116A423 for ; Mon, 6 Feb 2006 16:59:12 +0000 (GMT) (envelope-from adam@baylessfamily.org) Received: from bayless.fiber.net (bayless.fiber.net [216.83.146.126]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6CAC443D4C for ; Mon, 6 Feb 2006 16:59:12 +0000 (GMT) (envelope-from adam@baylessfamily.org) Received: from [10.1.102.220] (host-220.fbp.ore.fiber.net [216.83.157.220]) (authenticated bits=0) by bayless.fiber.net (8.13.3/8.13.3) with ESMTP id k16Gx8Oa076012 for ; Mon, 6 Feb 2006 09:59:12 -0700 (MST) (envelope-from adam@baylessfamily.org) Message-ID: <43E7806A.8000801@baylessfamily.org> Date: Mon, 06 Feb 2006 09:59:22 -0700 From: Adam Bayless User-Agent: Thunderbird 1.5 (Windows/20051201) MIME-Version: 1.0 To: freebsd-database@freebsd.org Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: MySQL 5 crashing on 6.0 X-BeenThere: freebsd-database@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Database use and development under FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 06 Feb 2006 16:59:13 -0000 I've got a brand new dual AMD opteron box with 4gigs total ram running MySQL 5.0.18. I built MySQL from ports and I'm using /usr/local/share/mysql/my-innodb-heavy-4G.cnf for my.cnf. The database is accessed by 4 client machines for an sql based spamassassin bayes database. The bayes tables are all InnoDB. The InnoDB data file is about 1.1 gigs on disk right now. MySQL is crashing and dumping a core file approximately every 4-6 hours. When it crashes mysqld_safe seems to be able to restart it just fine every time, but of course the innodb data has to be rebuilt from the logs which takes several minutes and processing mail stops while that happens. I thought initially that it must have been running into server memory limits, but I cut all of MySQL's internal limits approximately in half (in my.cnf) and it's still having the same trouble. There doesn't seem to be a certain query that causes it to crash every time, though when spamassassin runs a bayes expiration that sometimes seems to trigger the crash, other times it doesn't, though it always increases MySQL's CPU and memory usage. I've set the following in /boot/loader.conf: kern.maxdsiz="3684354560" kern.dfldsiz="3684354560" kern.maxssiz="3684354560" 'gdb -c mysqld.core' says: Program terminated with signal 11, Segmentation fault. #0 0x00000000dc9cf81f in ?? () which doesn't help me much. Maybe someone else can make some sense of that? Any ideas? Thanks, Adam