NFS server not responding

[Ryan Golhar]

I've google'd all over about this and can't seem to find a solution so
I'm hoping someone here can give some guidance...

I have a NFS server running Solaris in a far-off location

I have a bunch of RHEL 3 WS NFS clients mounting directories off the NFS
server in another location.

In /etc/fstab I have this entry to mount user home directories:

scifs:/users            /users                  nfs     rw
0 0

99% of the time things work well, but every once in a while the NFS
mount causes the machines to hang when used.

When I attempt to tar up a directory tree, the machine tends to hang.  I
wasn't sure what was going on at first but it looks like I have some
sort of an NFS issue.  

I keep seeing these messages in /var/log/messages:

Aug 25 15:36:54 kernel: nfs: server scifs not responding, still trying
Aug 25 15:39:09 kernel: nfs: server scifs not responding, still trying
Aug 25 15:39:09 kernel: nfs: server scifs OK Aug 25 15:40:09 kernel:
nfs: server scifs not responding, still trying Aug 25 15:40:32 kernel:
nfs: server scifs OK Aug 25 15:40:42 kernel: nfs: server scifs not
responding, still trying Aug 25 15:40:43 kernel: nfs: server scifs OK

I check with our network IT folks and they show minimal network usage
between the machines (they are on different subnets in different
locations).  I double-checked by doing an scp of an directory structure
from one machines local disk to another machines local disk and it is
fast, as expected.  

I ran ethereal to capture packets between the two machines and see the
tar process starting.  At some point, I start to see fragmented packets.
I'm not sure if this means anything (just yet).

So, I'm at a loss.  I don't know how to track down this problem, or even
where to start looking.  Google turned up similar issues, but there
isn't a consensus as to what is causing it.  Any ideas?

Ryan