Thanks for the response.
Very strange.... We added a battery to our RAID controller and then
switched our controller to write-back. ...after we booted into this
configuration 20-30% of our process were in this state and the machine
was S-L-U-G-G-I-S-H. We then rebooted to clear up the condition and it
reoccurred right away. We then switched the controller back to
write-through and the problem has gone away. We'll have 1-3 processes
in this state and they go away very fast, and the machine feels fine.
Do you have ANY clue what might cause this condition? We have an open
problem with Software support and they have analyzed our load and they
can't seem to find any problems.
--
jeffrey hundstad
alan_at_nabeth.cxo.dec.com wrote:
> Usually, a process in this state has entered the code for
> some device driver and the driver is waiting for a device
> to interrupt. In general it indicates that the whatever
> part of the kernel the process is using, is waiting for
> something to complete and it was unwise (or inconvient)
> to allow that wait to be interrupted. Whatever it is
> waiting for hasn't happened yet.
Received on Mon Feb 14 2000 - 21:07:51 NZDT