Hi,
Anton Ivanov wrote:
nfsd gets stuck in D state. Initially some machines, later all which read off
the nfs server fail to read. Messages like:
Mar 17 22:03:34 localhost kernel: [1899559.532028] statd: server rpc.statd not responding, timed out
Mar 17 22:03:34 localhost kernel: [1899559.532055] lockd: cannot monitor greebo
Mph, that's no good.
What kernel do you use now? Any changes? Can you use "alt+sysrq; w"
while in that state to get backtraces for blocked tasks, in case it
helps find a deadlock? (You may need to use "echo 1>/proc/sys/kernel/sysrq"
before that will work; see Documentation/sysrq.txt for details.)