On Wed, 10 Feb 1999 22:31:46 +0100 
 drochner@zel459.zel.kfa-juelich.de wrote:

 > I've noticed NFS swapping problems as well - there are situations
 > where a diskless machine simply doesn't recover from, in particular
 > when doing a huge compile (typilally C++) and the "scheduler - no room
 > for xxx" appears. Adding a local disk for swapping helped here.

I had this happen A LOT on a diskless Shark, and I made the problem
recoverable by making the pmap sleep for PT pages.

I.e. have the pagedaemon a chance to catch up and put pages on the free
list, and the system would recover.

 > Otoh, there might be general NFS problems too - it seems that a
 > diskless box doesn't survive an half-hour NFS server outage (v3 on UDP).
 > A lot of processes hang in "nfs_hashlock". (I think it was in "nfsrcvlk"
 > not so long ago.)

It's definitely true that there is a problem with "connected" NFS and
server outages.

        -- Jason R. Thorpe <thorpej@nas.nasa.gov>