NFS client loses shares
 

This has been a problem since I upgraded to Suse 9.0.

All of my computers share volumes with one another via NFS. After a certain period of time the NFS client one one or more of the computers will for some reason be unable to reach the exported shares of one or more of the other computers on the network. For example, computer A may no longer be able to see the shared volumes on computer B, although it can still see the shares on computer C (and both computer C and B can still see A). This is much more likely to occur if there are active users on both computers A and B, although it can still happen if just one of these computers has an active user (and it will happen, eventually, no matter what). If users on both computer A and computer B are writing to the hard drive of computer B at the same time, then it's a virtual certainty that A will lose its ability to see B within an hour, or even minutes.

Sometimes restarting the NFS client of computer A will fix the problem, at least until it drops out again. Sometimes this doesn't work and the only thing that seems to fix the problem is a reboot of computer A. Restarting both the NFS clients and servers of both A and B doesn't have the same effect as a reboot (i.e., it sometimes works, sometimes doesn't). A reboot of computer A will *always* fix the problem, without fail.

I'm assuming that I've set up either the NFS client or server for my machines incorrectly and that an insurmountable conflict of some sort occurs when both A and B are writing to the drive of computer B. I'm convinced it's a user problem (i.e., me) and not a software or hardware problem. I suspect I just need to add a simple option to fix this. Currently my NFS servers export their shares with: rw, no_root_squash and nothing more. My NFS clients import with: rw, bg, soft, retrans=50.

Any ideas?