From owner-freebsd-questions Thu Nov 21 18:49:49 2002 Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 5F84C37B401 for ; Thu, 21 Nov 2002 18:49:48 -0800 (PST) Received: from dan.emsphone.com (dan.emsphone.com [199.67.51.101]) by mx1.FreeBSD.org (Postfix) with ESMTP id D91DD43E91 for ; Thu, 21 Nov 2002 18:49:47 -0800 (PST) (envelope-from dan@dan.emsphone.com) Received: (from dan@localhost) by dan.emsphone.com (8.12.6/8.12.6) id gAM2ne6d040625; Thu, 21 Nov 2002 20:49:40 -0600 (CST) (envelope-from dan) Date: Thu, 21 Nov 2002 20:49:40 -0600 From: Dan Nelson To: David Smithson Cc: questions@FreeBSD.ORG Subject: Re: wanted: suggested sysctl variables for heavily used high-capacity file-server Message-ID: <20021122024940.GA19382@dan.emsphone.com> References: <011a01c291c1$4f186670$0801a8c0@customfilmeffects.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <011a01c291c1$4f186670$0801a8c0@customfilmeffects.com> X-OS: FreeBSD 5.0-CURRENT X-message-flag: Outlook Error User-Agent: Mutt/1.5.1i Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG In the last episode (Nov 21), David Smithson said: > Hi. I'm seeking suggestions for sysctl variables to use on a > highly-taxed mission-critical file-server. This computer exports two > 630GB arrays via SAMBA over gigabit on 3ware escalades. Anyone have > any tips? > > The reason I'm asking is because smbd has been dropping processes > lately. I recently added more render engines to our stack, so the > file-server seems to be choking somewhere. > > dmesg shows: > > -snip- > Nov 21 04:33:15 fnord /kernel: pid 1866 (smbd), uid 1005: exited on signal 6 > Nov 21 04:20:01 fnord /kernel: Limiting icmp unreach response from 208 to > 200 packets per second > -snip- Signal 6 is the abort signal, which means that smbd killed itself due to some unrecoverable error. If it's always smbd dying, check /var/log/log.smbd and /var/log/messages for possible errors, and if you can't find any, try running smbd on a vty (i.e. don't background it) and see if it prints anything to stderr when it dies. You might have found a bug in samba. The icmp message is simply a side-effect of smbd dying. You can ignore it. -- Dan Nelson dnelson@allantgroup.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message