Date: Wed, 8 Jul 2009 17:01:35 -0600 From: John Hein <jhein@timing.com> To: net@freebsd.org Subject: Re: network lock manager (lockd) deadlocked in 'rpcrecv' Message-ID: <19029.9551.628427.146587@gromit.timing.com> In-Reply-To: <19029.5367.534192.928426@gromit.timing.com> References: <19029.4145.296260.915327@gromit.timing.com> <19029.5367.534192.928426@gromit.timing.com>
next in thread | previous in thread | raw e-mail | index | archive | help
John Hein wrote at 15:51 -0600 on Jul 8, 2009: > John Hein wrote at 15:31 -0600 on Jul 8, 2009: > > I have a home directory on FreeBSD 7.2-stable (20090705), amd64. > > It is serving up the directory over nfs (v3, tcp), and now > > I'm seeing lots of 'lockd not responding' on Fedora 10 & 11 systems. . . > Also in dmesg: > > NLM: failed to contact remote rpcbind, stat = 5, port = 28416 . . Here's some good information. This seems to happen when there are 2 or more Fedora systems trying to access locks via lockd. Rebooting the Fedora box that has 'lockd not responding' frees up the deadlocked freebsd lockd. But _also_ disabling the firewall on the Fedora boxes helps, too. This doesn't necessarily completely implicate or exonerate lockd. But what should lockd do when the remote box asks for a lock, but doesn't complete the RPC dialog? Is there a way we can deal with this problem and not have lockd deadlock?
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19029.9551.628427.146587>