Subject: Re: Some more 'D' evidence.
To: Darren Reed <darrenr@vitruvius.arbld.unimelb.edu.au>
From: David Gilbert <dgilbert@jaywon.pci.on.ca>
List: port-sparc
Date: 04/07/1996 14:54:17
>>>>> "Darren" == Darren Reed <darrenr@vitruvius.arbld.unimelb.edu.au> writes:

Darren> In some email I received from David Gilbert, sie wrote:
>>  A number of people have reported unkillable processes stuck in the
>> 'D' (Uninteruptable Disk Wait) state.  I have a bit more evidence
>> for the fire.
>> 
>> I have found that if I get a process stuck in 'D', then I can
>> produce more such processes by attempting to read the directory in
>> which that process is attempting to read files.  I have been able
>> to duplicate this where I had compiles running which I could
>> definately track down.
>> 
>> Interestingly, nfsd processes will also get stuck in this manner.
>> I don't know if this is interesting, or not.

Darren> This is just NFS losing.  Probably means your network has
Darren> fallen over (or the NFS server) and you don't have the
Darren> partition mounted with "intr" or "soft" options.

	How so?  In a recent case, the nfsd processes got stuck in 'D'
and listing the directory that they were working on stuck the 'ls'
process in 'D' as well.

	How can I be sure what directory is being read and that my
network is not 'falling down' ??? I only have two machines on the
net in question.

Dave.

-- 
----------------------------------------------------------------------------
|David Gilbert, PCI, Richmond Hill, Ontario.  | Two things can only be     |
|Mail:      dgilbert@jaywon.pci.on.ca         |  equal if and only if they |
|http://www.pci.on.ca/~dgilbert               |   are precisely opposite.  |
---------------------------------------------------------GLO----------------