From owner-freebsd-current@FreeBSD.ORG Fri Jun 11 18:11:22 2004 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 1FC7F16A4CE for ; Fri, 11 Jun 2004 18:11:22 +0000 (GMT) Received: from freefall.freebsd.org (freefall.freebsd.org [216.136.204.21]) by mx1.FreeBSD.org (Postfix) with ESMTP id 1797E43D1D; Fri, 11 Jun 2004 18:11:22 +0000 (GMT) (envelope-from bmilekic@FreeBSD.org) Received: from freefall.freebsd.org (bmilekic@localhost [127.0.0.1]) i5BIB5Du029092; Fri, 11 Jun 2004 18:11:05 GMT (envelope-from bmilekic@freefall.freebsd.org) Received: (from bmilekic@localhost) by freefall.freebsd.org (8.12.11/8.12.11/Submit) id i5BIB2sO029091; Fri, 11 Jun 2004 18:11:02 GMT (envelope-from bmilekic) Date: Fri, 11 Jun 2004 18:11:02 +0000 From: Bosko Milekic To: othermark Message-ID: <20040611181102.GA25027@freefall.freebsd.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.4.1i cc: freebsd-current@freebsd.org Subject: Re: Today's -current panics X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 11 Jun 2004 18:11:22 -0000 othermark wrote: >It's all a default installation (I try to keep the machines as untweaked as >possible). The only thing in loader.conf is to turn on DMA for the hdd. > >I rebooted with kern.ipc.nmbclusters="0" and vm.kmem_size="300000000" in >/boot/loader.conf. > >Another NFS installworld completed successfully (but appeared to going >slower). I'm assuming that GENERIC w/ no adjustments should complete this >type of task w/o panicing anyway? Thanks for your help Bosko! Be careful, though, if you are running GENERIC here. GENERIC by default includes debugging options such as INVARIANTS and WITNESS which will significantly affect performance, so that could be why it feels slower, in particular if you were not running with those options before. Also be careful if you ever intend to further increase vm.kmem_size, as it might also involve having to increase KVA_PAGES (kernel KVA). On another note, I noticed from your 'netstat -m' output that you have many mbufs allocated. If you reduce/stop the NFS load, does the number of allocated mbufs ever drop? If not, you could be hitting an mbuf leak somewhere (possibly in the NFS code). Also check to make sure that the debug.mpsafenet sysctl is set to zero (it should be by default, unless you're toggling it explicitly), just to be on the safe side. -Bosko