Re: [NFS] processes stuck in D state

Re: [NFS] processes stuck in D state

[Michael Buesch]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Tuesday 06 May 2003 17:20, Trond Myklebust wrote:
> >>>>> " " == Michael Buesch <fsdeveloper@yahoo.de> writes:
>      > Hi!  Please take a look at this problem:
>      >
>      > [linux-kernel-mailing-list thread]
>      > http://marc.theaimsgroup.com/?t=98639966100003&r=1&w=2
>
> If I can hazard a guess: someone is firewalling the lockd port and/or
> the statd port.
>
> Either mount using the 'nolock' option, or fix the firewall (see the
> HOWTO and/or FAQ).

To reproduce the problem:
- - mount some nfs from a server in your lan.
- - Open an app, that uses the mounted fs. I've simply opened a
  konqueror-window for the directory where the nfs is mounted.
- - shut down or crash the server or just pull the network-cable.
- - Now the konqueror-process is nonkillable in D state. There's no
  chance to kill it.

I've tried it with all firewalls disabled, but the problem resists.

> Cheers,
>   Trond

@linux-kernel-mailing-list: I've posted a thread to nfs-mailing list with
the same topic as in lkml. IMHO this is the better list for this problem. :)

- -- 
Regards Michael Büsch
http://www.8ung.at/tuxsoft
 17:34:35 up  1:26,  5 users,  load average: 1.52, 1.32, 1.13
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQE+t9fWoxoigfggmSgRAvrcAJ4i+i3V+kcRd+kLHS7cb2WDZDHKsQCfWljd
rwtAFK4ONkJHzVck03t7F5U=
=gHuP
-----END PGP SIGNATURE-----

Re: [NFS] processes stuck in D state

[Trond Myklebust]

>>>>> " " == Michael Buesch <fsdeveloper@yahoo.de> writes:


     > To reproduce the problem:
     > - - mount some nfs from a server in your lan.
     > - - Open an app, that uses the mounted fs. I've simply opened a
     >   konqueror-window for the directory where the nfs is mounted.
     > - - shut down or crash the server or just pull the
     >     network-cable.
     > - - Now the konqueror-process is nonkillable in D
     >     state. There's no
     >   chance to kill it.

Unless you are using the 'intr' or 'soft' mount flags, then that is
*documented and expected* behaviour.

It is true that even when using the 'intr' mount flag, you don't
always succeed in killing a task that is hanging on NFS. That is
usually due to the fact that it is waiting on some semaphore that is
held by another process. semaphores always sleep in the
TASK_UNINTERRUPTIBLE state, so they cannot be signalled.
Linus has suggested a solution to this problem: to set up a special
class of semaphores that are killable with 'SIGKILL', but doing that
(and then replacing all those semaphores in the VFS and VM) is not
going to happen before 2.7.x. at the earliest.

However, as I've mentioned on this list *many* times before: there
exists a workaround if you are wanting to kill all processes in order
to unmount the partition:
  kill -9 all the processes.
  kill -9 rpciod.

Cheers,
  Trond

Re: [NFS] processes stuck in D state

[Michael Buesch]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Tuesday 06 May 2003 18:05, Trond Myklebust wrote:
> >>>>> " " == Michael Buesch <fsdeveloper@yahoo.de> writes:
>      > To reproduce the problem:
>      > - - mount some nfs from a server in your lan.
>      > - - Open an app, that uses the mounted fs. I've simply opened a
>      >   konqueror-window for the directory where the nfs is mounted.
>      > - - shut down or crash the server or just pull the
>      >     network-cable.
>      > - - Now the konqueror-process is nonkillable in D
>      >     state. There's no
>      >   chance to kill it.
>
> Unless you are using the 'intr' or 'soft' mount flags, then that is
> *documented and expected* behaviour.

I'm using intr.

> However, as I've mentioned on this list *many* times before: there
> exists a workaround if you are wanting to kill all processes in order
> to unmount the partition:
>   kill -9 all the processes.
>   kill -9 rpciod.

kill -9 doesn't work for me to kill the app.

- -- 
Regards Michael Büsch
http://www.8ung.at/tuxsoft
 18:28:55 up  2:20,  5 users,  load average: 1.02, 1.06, 1.06
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQE+t+MhoxoigfggmSgRAkeqAJ0c71DxLZ13/CHqUXlTa8TvjAt2iwCeLO34
s7crt56Gr8JyKxCLZMbrNvc=
=z8EU
-----END PGP SIGNATURE-----

Re: [NFS] processes stuck in D state

[Trond Myklebust]

>>>>> " " == Michael Buesch <fsdeveloper@yahoo.de> writes:

    >> kill -9 all the processes.  kill -9 rpciod.

     > kill -9 doesn't work for me to kill the app.

I didn't say kill the app. I said signal it with -9, then signal
rpciod.

Cheers,
  Trond

Re: [NFS] processes stuck in D state

[Michael Buesch]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Tuesday 06 May 2003 18:54, Trond Myklebust wrote:
> >>>>> " " == Michael Buesch <fsdeveloper@yahoo.de> writes:
>     >> kill -9 all the processes.  kill -9 rpciod.
>     >>
>      > kill -9 doesn't work for me to kill the app.
>
> I didn't say kill the app. I said signal it with -9, then signal
> rpciod.

Ah, I understand. :)

> Cheers,
>   Trond

- -- 
Regards Michael Büsch
http://www.8ung.at/tuxsoft
 19:31:20 up  3:22,  2 users,  load average: 1.23, 1.09, 1.04
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQE+t/G3oxoigfggmSgRAq5BAJ0SezM+y1LFnwglArReHERXb2VJZQCeKKd0
Sx6RqCkOvm4FvgTCVyx2gCE=
=K8c7
-----END PGP SIGNATURE-----

RE: [NFS] processes stuck in D state

[Lever, Charles]

> To reproduce the problem:
> - - mount some nfs from a server in your lan.
> - - Open an app, that uses the mounted fs. I've simply opened a
>   konqueror-window for the directory where the nfs is mounted.
> - - shut down or crash the server or just pull the network-cable.
> - - Now the konqueror-process is nonkillable in D state. There's no
>   chance to kill it.

does the problem persist after you reconnect the network cable?
what happens when the server becomes available again?
are you mounting with UDP or TCP?

Re: [NFS] processes stuck in D state

[Michael Buesch]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Tuesday 06 May 2003 17:47, Lever, Charles wrote:
> does the problem persist after you reconnect the network cable?
> what happens when the server becomes available again?

no. If server is available again, the process wakes up from D.

But like man mount says:
[snip] The process cannot be interrupted or killed unless you also specify intr. [/snip]
The process should be killable while the cable is pulled.
But that's not the case, although intr is in fstab.

> are you mounting with UDP or TCP?

uh. How to find it out? :)

- -- 
Regards Michael Büsch
http://www.8ung.at/tuxsoft
 17:53:19 up  1:44,  5 users,  load average: 1.04, 1.05, 1.06
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQE+t9s0oxoigfggmSgRAkDPAKCFMeEGvS3KUhwn0bNQngKRK6h2fwCdEcv/
U2ttfZ6Mm8Sazuksfn4UUrY=
=M5Kh
-----END PGP SIGNATURE-----