OES LINUX SP1 cluster
shared storage, with nss pools/vols

I currently have around 200-300 users connected to the shared storage,
with novell client 4.9sp2
works fine
except, when the resource (nss pool/vol) is for too long on 1 server, it
stops responding, and i have to manually migrate it to the other server,
and restart the other one.
It's like if ncpserv stops reponding, for no reason
Usually, max 24h.

Any clue ?
Or is this another OES Linux bug ?