[Linux-cluster] GFS + DRBD Problems

[Gordan Bobic]

Bryn M. Reeves wrote:

>>> I have a 2-node cluster with Open Shared Root on GFS on DRBD. A single
>>> node mounts GFS OK and works, but after a while seems to just block
>>> for disk.
>> [...]
>> State:  D (disk sleep)
>> SleepAVG:       97%
>> [...]
> 
> You can find out what it's sleeping on either by via a sysrq or by
> getting ps to display the wchan field of the stat data, e.g.:
> 
> ps ac -opid,comm,wchan
> 
> And see what symbol appears in the 3rd field.

And this is what comes out for all the stuck processes:

ps ax -opid,comm,wchan | grep ssh
  9250 sshd            -
  9507 sshd            gdlm_plock
  9642 sshd            gdlm_plock

They are all stuck in the gdlm_plock function. I figured it'd be 
something like this. How do I debug this further?

Gordan