From owner-freebsd-afs@FreeBSD.ORG Wed Apr 14 16:52:20 2010 Return-Path: Delivered-To: freebsd-afs@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 78E331065672 for ; Wed, 14 Apr 2010 16:52:20 +0000 (UTC) (envelope-from kaduk@mit.edu) Received: from dmz-mailsec-scanner-5.mit.edu (DMZ-MAILSEC-SCANNER-5.MIT.EDU [18.7.68.34]) by mx1.freebsd.org (Postfix) with ESMTP id 319C68FC22 for ; Wed, 14 Apr 2010 16:52:19 +0000 (UTC) X-AuditID: 12074422-b7c13ae000003829-55-4bc5f2c38e8b Received: from mailhub-auth-2.mit.edu (MAILHUB-AUTH-2.MIT.EDU [18.7.62.36]) by dmz-mailsec-scanner-5.mit.edu (Symantec Brightmail Gateway) with SMTP id 7F.DC.14377.3C2F5CB4; Wed, 14 Apr 2010 12:52:19 -0400 (EDT) Received: from outgoing.mit.edu (OUTGOING-AUTH.MIT.EDU [18.7.22.103]) by mailhub-auth-2.mit.edu (8.13.8/8.9.2) with ESMTP id o3EGqIXK028380; Wed, 14 Apr 2010 12:52:19 -0400 Received: from multics.mit.edu (MULTICS.MIT.EDU [18.187.1.73]) (authenticated bits=56) (User authenticated as kaduk@ATHENA.MIT.EDU) by outgoing.mit.edu (8.13.6/8.12.4) with ESMTP id o3EGqGHg028790 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NOT); Wed, 14 Apr 2010 12:52:17 -0400 (EDT) Received: (from kaduk@localhost) by multics.mit.edu (8.12.9.20060308) id o3EGqFvK001278; Wed, 14 Apr 2010 12:52:15 -0400 (EDT) Date: Wed, 14 Apr 2010 12:52:15 -0400 (EDT) From: Benjamin Kaduk To: Jan Henrik Sylvester In-Reply-To: <4BC5EEB5.5050109@janh.de> Message-ID: References: <558186334.2619.1271175586732.JavaMail.root@thunderbeast.private.linuxbox.com> <4BC4A5F1.6050406@janh.de> <4BC5EEB5.5050109@janh.de> User-Agent: Alpine 1.10 (GSO 962 2008-03-14) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; format=flowed; charset=US-ASCII X-Brightmail-Tracker: AAAAAA== Cc: afs-list freebsd Subject: Re: AFS on FreeBSD 8? X-BeenThere: freebsd-afs@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: The Andrew File System and FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 14 Apr 2010 16:52:20 -0000 On Wed, 14 Apr 2010, Jan Henrik Sylvester wrote: > On 04/13/2010 19:20, Benjamin Kaduk wrote: >> On Tue, 13 Apr 2010, Jan Henrik Sylvester wrote: > > Sometimes access to afs takes a long time, for example the first 'ls' for a > server after 'cd /afs/SOMESERVER' takes a few minutes. The next few commands > only took a minute each, until it was at reasonable speed. > > Copying down a large file from secured afs worked pretty well (with about > half of the possible 100mbit bandwidth). The CPU system load was pretty high, > though: 12% on a C2Q. > > Then I tried to copy a file to afs... since then, afs hangs (even 'cd /afs' > in a different terminal hangs). Fascinating. I have been seeing almost precisely the reverse behavior -- I could write files into AFS without difficulty, but any attempt to read from AFS would cause the process to hang unkillably. While that was hung, other processes could traverse (at least parts of) /afs, though. > > Even if nothing uses afs, 'umount /afs' hangs, too, but I can still issue > 'afsd -shutdown' in a different terminal after that. Restarting afsd freezes > the machine, though. > > Overall, afs does not seem to be usable, but you did not promise anything > else. > > Is there any way to cleanly shut down afs and restart it for further tests > without reboot the machine? > > Where can I get the rc script Derrick was talking about? > I am given to understand that the rc script unloads and reloads the kernel module before restarting afsd; it should be in ${WRKSRC}/src/afsd/afs.rc.fbsd . That said, I haven't looked at it, and have just been rebooting my machine between tests. > Any ideas for improvements? Should 8-STALBE instead of 8.0-RELEASE work > better? Any configuration changes I should try? I don't expect -STABLE vs. -RELEASE to have much effect; I'm pretty sure these are openafs bugs. Could you send a bit of detail about your machine, though? Is it single-processor, 32-bit, etc.?