From owner-freebsd-current@FreeBSD.ORG Fri Jan 8 19:58:31 2010 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 016BF1065670; Fri, 8 Jan 2010 19:58:31 +0000 (UTC) (envelope-from ltning@anduin.net) Received: from mail.anduin.net (mail.anduin.net [213.225.74.249]) by mx1.freebsd.org (Postfix) with ESMTP id A9F608FC0A; Fri, 8 Jan 2010 19:58:30 +0000 (UTC) Received: from [212.62.248.147] (helo=ranger.home.anduin.net) by mail.anduin.net with esmtpsa (TLSv1:AES128-SHA:128) (Exim 4.69 (FreeBSD)) (envelope-from ) id 1NTKyZ-000Lj0-4G; Fri, 08 Jan 2010 20:58:27 +0100 Mime-Version: 1.0 (Apple Message framework v1077) Content-Type: text/plain; charset=iso-8859-1 From: =?iso-8859-1?Q?Eirik_=D8verby?= In-Reply-To: <3BC13C71-348A-49AD-94BA-FC28EB0648DE@freebsd.org> Date: Fri, 8 Jan 2010 20:58:26 +0100 Content-Transfer-Encoding: quoted-printable Message-Id: <43D0E400-183E-4741-9ABD-14ED80D90983@anduin.net> References: <20091129013026.GA1355@michelle.cdnetworks.com> <74BFE523-4BB3-4748-98BA-71FBD9829CD5@anduin.net> <34AD565D-814A-446A-B9CA-AC16DD762E1B@anduin.net> <1DFC4992-E136-4674-BC0E-A6B1DAE12AF4@anduin.net> <3BC13C71-348A-49AD-94BA-FC28EB0648DE@freebsd.org> To: "Robert N. M. Watson" , Rick Macklem X-Mailer: Apple Mail (2.1077) Cc: pyunyh@gmail.com, weldon@excelsusphoto.com, freebsd-current@freebsd.org, Gavin Atkinson Subject: Re: FreeBSD 8.0 - network stack crashes? X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 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, 08 Jan 2010 19:58:31 -0000 On 8. jan. 2010, at 16.33, Robert N. M. Watson wrote: > On 30 Nov 2009, at 19:13, Eirik =D8verby wrote: >=20 >> I meant NFS-UDP ... However I was wrong even there; Using NFS over = UDP from FreeBSD boxes does not cause the same issue. So OpenBSD seems = to be a special case here. >>=20 >> I'm no Wireshark expert (to be fair, I've seen it a few times and = tried it once or twice, and that's so long ago it's almost no longer = true), so I'd need some input on how to gather useful data. I assume = tcpdump, which options? And would it be OK if I made the dump available = for download somewhere, so you or someone else can take a look with = whichever tools you'd like? >=20 > Aii. Over a month zips past in the blink of an eye. Happens all the time. ;) > Are you still experiencing this problem? I can certainly look at a = wireshark trace, but make no promises. If you do do a trace, then what = we should do is have you do run a script that dumps a bunch of relevant = stats with nfsstat, netstat, vmstat, etc, before the trace starts, grabs = exactly ${someval} seconds of trace data, then dumps all the same stats = afterwards. Then we can use the stats to work out about how many leaked = packets (or whatever) were present, and try to correlate it to a count = of some type of event in the trace. I got a patch from Rick early december which I have tried, but that one = unfortunately did not seem to make any difference. I cvsup'ed = (RELENG_8_0), patched, compiled, installed, booted and tested on the = 12th - any reason to think things might have changes on 8.0 since then? = I'm assuming not, in which case the answer is yes, I am still = experiencing this problem (except I've made all our OpenBSD boxen speak = TCP now, avoiding the problem entirely). I'll perform whichever tests you'd like me to on that particular system. = Just send me the scripts or pseudocode, and I'll get it done as soon as = I can. I was also planning on setting up a pair of VMs (one FreeBSD and one = OpenBSD) and try to reproduce there - but that month that zipped past = has not allowed the time to do so. Thanks, /Eirik