From owner-freebsd-hackers Mon Apr 14 08:18:33 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA20119 for hackers-outgoing; Mon, 14 Apr 1997 08:18:33 -0700 (PDT) Received: from david.siemens.de (david.siemens.de [139.23.36.11]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id IAA20114 for ; Mon, 14 Apr 1997 08:18:30 -0700 (PDT) Received: from salomon.mchp.siemens.de (salomon.siemens.de [139.23.33.13]) by david.siemens.de (8.8.5/8.8.5) with ESMTP id RAA24308 for ; Mon, 14 Apr 1997 17:13:29 +0200 (MDT) Received: from curry.mchp.siemens.de (1@curry.mchp.siemens.de [146.180.31.23]) by salomon.mchp.siemens.de (8.8.4/8.8.4) with ESMTP id RAA06935 for ; Mon, 14 Apr 1997 17:18:21 +0200 (MDT) Received: (from daemon@localhost) by curry.mchp.siemens.de (8.8.5/8.8.5) id RAA02763 for ; Mon, 14 Apr 1997 17:18:19 +0200 (MET DST) From: Andre Albsmeier Message-Id: <199704141518.RAA06336@curry.mchp.siemens.de> Subject: Re: mbuf clusters problem in 2.2.1R In-Reply-To: <199704111115.PAA11170@sinbin.demos.su> from "Alex G. Bulushev" at "Apr 11, 97 03:15:42 pm" To: bag@sinbin.demos.su (Alex G. Bulushev) Date: Mon, 14 Apr 1997 17:18:09 +0200 (CEST) Cc: hackers@freebsd.org X-Mailer: ELM [version 2.4ME+ PL31 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-hackers@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > we use p6-200 with two 3940UW and fxp (100baseT fullduplex) 256M RAM > under fbsd 2.2.1R > > system reboots once per 2-6 days with out of mbuf clusters diagnostic > on console ... we use maxusers=256 without any effect, then we > increase NMBCLUSTERS to 8192 and then to 12288 with the same effect ... > > while working, number of allocated mbuf clusters permanently increase > (some mbufs not freeing?) and then system crashed ... > > this effect exists only on system with big network traffic ... > now we run: > squid - up to 200 connection simultaneously > ircd - up to 50 connection > cvsupd ... I have got similar weird things.i I often see this in the logfiles: Apr 11 13:07:28 server kern.info /kernel: nfsd send error 55 Apr 11 13:07:28 server last message repeated 719 times Apr 11 13:07:28 server kern.err /kernel: Out of mbuf clusters - increase maxusers! Apr 11 13:07:28 server kern.info /kernel: nfsd send error 55 Apr 11 13:07:28 server last message repeated 21 times Apr 11 13:07:28 server kern.crit /kernel: 55 Apr 11 13:07:28 server kern.info /kernel: nfsd send error 55 Apr 11 13:07:28 server last message repeated 743 times Apr 11 13:07:29 server kern.crit /kernel: 55 Apr 11 13:07:29 server kern.info /kernel: nfsd send error 55 Apr 11 13:07:29 server last message repeated 743 times >From time to time the machine reboots without any other messages or console errors. This only appears on a machine with 3 network cards and heavy net traffic. What I found curiuos is the line: /kernel: 55. I have changed kernel options, network cards, memory, motherboard, ... no change. This is a 2.2.1 system which serves a lot of PCs. During the night everything is sane but when the PCs are running the dmesg fills with "nfsd send error 55" and occasionaly crashes occur. I have no clue where to start or what to change... -Andre