From owner-freebsd-stable Mon Jul 24 8:36:11 2000 Delivered-To: freebsd-stable@freebsd.org Received: from ahab.com (gw1-75ml.themediatc.com [204.143.179.2]) by hub.freebsd.org (Postfix) with ESMTP id 915AE37B90E for ; Mon, 24 Jul 2000 08:36:03 -0700 (PDT) (envelope-from moxie@ahab.com) Received: (from moxie@localhost) by ahab.com (8.9.3/8.9.3) id QAA00551; Fri, 21 Jul 2000 16:36:15 -0400 (EDT) (envelope-from moxie) From: Jason T Date: Fri, 21 Jul 2000 16:36:15 -0400 To: Philippe Le Berre Cc: Noor Dawod , "Chris D. Faulhaber" , freebsd-stable@FreeBSD.ORG Subject: Re: 4.1 RC locked-up with MySql 3.22.32 and 3.23.21b Message-ID: <20000721163614.A283@sseye.ahab.com> Mail-Followup-To: Philippe Le Berre , Noor Dawod , "Chris D. Faulhaber" , freebsd-stable@FreeBSD.ORG References: <4.3.2.7.2.20000720122927.00c32cd0@212.129.1.43> <4.3.2.7.2.20000721081753.00c4bbe0@212.129.1.43> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.4i In-Reply-To: <4.3.2.7.2.20000721081753.00c4bbe0@212.129.1.43>; from philippe@le-berre.com on Fri, Jul 21, 2000 at 08:22:51AM -0700 Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Hmm... I'm still not convinced... I have used MySQL from .22 up to .32 and I have experienced periodic (about once a month, strangely) signal 11s on all of them, both on solaris and freebsd. I finally resorted to restarting mysql in the dead of night, and since then (.32, freebsd-3.4-stable), it's been just fine... The crashes always happened under low load in the wee hours. Not like I tried a debugger or anything... On Fri, Jul 21, 2000 at 08:22:51AM -0700, Philippe Le Berre wrote: > Hi, > > Thanks for the info, I eventually found out the issue. I had the server > kernel build with maxuser at 64, reducing that to 32 has somehow disable > the 88% cpu limitation on the process and therefore I was able to > successfully run the full MySQL test suite. However, how comes that on > FreeBSD the MySQL test suite pushes so hard on CPU? I have never seen that > with other O/S. And the code that kills a process reaching the CPU limits > is buggy because it does provoke a server locked-up (no even a panic, just > a big cold freeze on the box). > > Regards, > > -philippe > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-stable" in the body of the message -- ------------------------- Jason Thaxter Flashlight Media jason@flashlightmedia.com www.flashlightmedia.com ------------------------- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message