Re: System hangs due to NFS share

[Sylvain <sylvaintersideral@gmail.com>]

2012/1/23 Frank <frank@anotheria.net>:
> On Mon, 2012-01-23 at 11:47 +0100, Sylvain wrote:
>
>> I already tried this option too, but it didn't help. Also according to
>> the manpage:
>>
>> > The intr / nointr mount option is deprecated after kernel 2.6.25.  Only SIGKILL can interrupt a pending NFS operation on these kernels, and  if  specified, this mount option is ignored to provide backwards compatibility with older kernels.
>
> OK, the man page is not yet available on lenny.
>
> Have you tried to kill the process using SIGKILL? What happens if you
> do? I assume this behaviour is related to the state TASK_KILLABLE which
> was introduced in kernel 2.6.25. In previous kernels a process waiting
> for I/O was in TASK_UNINTERRUPTIBLE and never get the signal you send
> it.
>
> Frank

I tried to SIGKILL the [nfsiod] process but it didn't get killed. I
also tried to kill the various rpcbind processes but it didn't change
anything.

I also tried autofs as Scott suggested, but the same problem occured
when I unplugged the network cable (most of the system hangs, can't
reboot nor halt). Here's how I configured autofs:

/etc/auto.misc:
bazinga -retry=1,rw,hard,size=8192,wsize=8192 192.168.2.77:/DataVolume/Public

/etc/auto.master:
/misc   /etc/auto.misc --timeout 20

Here's what I found in the syslog:

> Jan 23 22:59:17 cid kernel: [  510.944160] nfs: server 192.168.2.77 not responding, still trying
> Jan 23 23:01:22 cid kernel: [  635.616161] nfs: server 192.168.2.77 not responding, still trying
> Jan 23 23:01:46 cid sm-notify[841]: DNS resolution of Bazinga.local failed; retrying later
> Jan 23 23:03:51 cid sm-notify[841]: DNS resolution of Bazinga.local failed; retrying later
> Jan 23 23:05:56 cid sm-notify[841]: DNS resolution of Bazinga.local failed; retrying later
> Jan 23 23:07:56 cid sm-notify[841]: Unable to notify Bazinga.local, giving up

I'm not sure why it's trying to resolve the "bazinga.local" name, and
even when it gave up with the resolution, it didn't unfreeze anything.
Also nfs seems to be still trying to reach the server (the cable was
unplugged at 22:56).

Thanks for your help,
Sylvain