From owner-freebsd-net@FreeBSD.ORG Mon Aug 15 20:46:29 2011 Return-Path: Delivered-To: freebsd-net@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id B20F7106566C for ; Mon, 15 Aug 2011 20:46:29 +0000 (UTC) (envelope-from sebas@koumbit.org) Received: from bureau.koumbit.net (homere.koumbit.net [209.44.112.81]) by mx1.freebsd.org (Postfix) with ESMTP id 8A1458FC15 for ; Mon, 15 Aug 2011 20:46:29 +0000 (UTC) Received: from [192.168.20.52] (modemcable012.160-130-66.mc.videotron.ca [66.130.160.12]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by bureau.koumbit.net (Postfix) with ESMTPSA id 7DCB6E80E84 for ; Mon, 15 Aug 2011 16:29:30 -0400 (EDT) Message-ID: <4E4981AA.2010907@koumbit.org> Date: Mon, 15 Aug 2011 16:29:30 -0400 From: =?ISO-8859-1?Q?S=E9bastien_Grenier?= User-Agent: Lightningcow/5.0 MIME-Version: 1.0 To: freebsd-net@freebsd.org X-Enigmail-Version: 1.2 X-Mailer: Camino/Lightningcow Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Subject: ping "No buffer space available" (freebsd 8.2) X-BeenThere: freebsd-net@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Networking and TCP/IP with FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 15 Aug 2011 20:46:29 -0000 Hi folks We have a Sun Server X2200 M2 machine[1] running FreeBSD 8.2, and we are having serious reliability issues with the network routing. This router used to be another machine (a generic PC) running OpenBSD 4.2, and was running perfectly fine. After a while (a few hours to a few days), we get the following error message when trying to ping any IP: "No buffer space available". After doing some research, we have tried to do some tuning of the following kernel parameters: sysctl -w net.inet.tcp.sendbuf_max = 16777216 sysctl -w net.inet.tcp.recvbuf_max = 16777216 sysctl -w kern.ipc.maxsockbuf=400000 sysctl -w net.inet.tcp.recvspace=512000 We still get the error message "No buffer space available" after a while. So far, the only wait to get around the problem is to reboot the machine and reapply the above configuration, which keeps the problem away for a few days instead of a few hours. We've enabled the collection of statistics on the router with munin[2] We didn't apply the above configurations in /boot/loader.conf because we still experience the error... Any idea? What would be a good way to debug? Are they values that we should be monitoring? Regards, Sébastien Koumbit.org [1] more details here: https://wiki.koumbit.net/rtr0.koumbit.net [2] http://stats.koumbit.net/koumbit.net/rtr0.koumbit.net/index.html (when it's routing ok.) rtr0# netstat -m 829/2246/3075 mbufs in use (current/cache/total) 799/1861/2660/51200 mbuf clusters in use (current/cache/total/max) 798/1378 mbuf+clusters out of packet secondary zone in use (current/cache) 0/35/35/25600 4k (page size) jumbo clusters in use (current/cache/total/max) 0/0/0/12800 9k jumbo clusters in use (current/cache/total/max) 0/0/0/6400 16k jumbo clusters in use (current/cache/total/max) 1807K/4423K/6230K bytes allocated to network (current/cache/total) 0/0/0 requests for mbufs denied (mbufs/clusters/mbuf+clusters) 0/0/0 requests for jumbo clusters denied (4k/9k/16k) 0/0/0 sfbufs in use (current/peak/max) 0 requests for sfbufs denied 0 requests for sfbufs delayed 0 requests for I/O initiated by sendfile 0 calls to protocol drain routines -- L'éducation n'est pas une marchandise, L'éducation est un droit Luttons pour ce droit http://www.youtube.com/watch?v=ek2B3Nd7Ofk