Hi! I am running a Novell 3.12 Server (25 users) with NFS module 1.27b and
TCP/IP module loaded. I have 2 Linux 1.2.11 hosts connecting to the Novell
Server and from the 2 Linux hosts, I can use NFS to mount volumes on the
Novell Server without any problem.
I am running some applications on the 2 Linux hosts accessing the Novell
file systems mounted via NFS for more than 3 years with no NFS errors.
But recently (I did not add or modify my applications on Linux or touch
the Novell server) the Novell server reports large amount of error
message on PKERNEL.NLM, and the error says:
PKERNEL: Warning: RPC/UDP receive queue are full, packet dropping
This error message appears 3 to 6 times per second and flooded the screen.
The Novell server also fails to accept new connections and all existing
connections (including NFS) are also stalled. When I did a "down" on
the system console, it hangs for a long time and I have to cold boot.
I have gone thr' all the manuals and FAQs and cannot find any related
cure for the above problem, but I think it is related to NFS, because
when I shutdown the 2 Linux hosts, the error message stops.
Can anyone direct me or give me some clue on how to avoid this problem?