NFS Server Questions

Brad Loomis (
Mon, 17 Nov 1997 19:10:16 -0500


I am experiencing some interesting problems with a medium size Linux compute
cluster and could use some insight. Here is some background info.

29 Linux compute servers running Redhat 4.1 with Kernel 2.0.31 on SCSI
drives. 256M of RAM with 1GB of swap. All identical with 3comm Boomerang
NIC's (100Mbs). We are also using amd and here is the entry in the
/etc/amd.conf for the nfs server options that we push out to the compute


2 Linux NFS servers running Redhat 4.1 with kernel 2.0.31, again 3comm
Boomerang NIC's with Full Duplex. These are data stores for the compute
servers to mount, read in the data and push the corresponding results back out.

1. Cabletron MMAC Plus with Fast Ethernet Cards connecting this all together
in one big flat net.

Reading several FAQ's, and HOWTO's I understand the Linux does not provide
read and write caching for nfs mounts. Thus preformance over nfs related
mounts can be degraded. However, I am seeing an interesting problem when a
large amount of clients are initially requesting reads from the NFS server,
several of the other clients (at random) hang on existing mount points to
the same nfs server. I can make new mounts to other exported partitions on
the nfs server, but the any existing mounts will hang and may not recover.
The load on the nfs server is minimal and I am seeing no network errors on
the clients or the server via netstat.

Not knowing of a nfswatch or nfstat program for linux, I am kind of limited
in my troubleshooting of this problem.

Has anyone else seen this and know of possible solutions?? Does anyone know
of a linux tool to debug an NFS related problems?? Any
help/pointers/suggestions would be greatly appreciated.

As always thanx in advance,

Brad Loomis
Network Engineer Phone:781-937-3338 ext329
Net Daemons Assoc., 800 W. Cummings Park Drive,#2050 Woburn MA 01801
"It is kind of fun to do the impossible --Walt Disney"