We have a NFS server (running RHEL4) to support a large number of client workstations (running RHEL4). The clients are all identical (in term of hardware, OS, system configuration and etc), and have /usr/local mounted from the NFS server. /etc/fstab is configured identically on all clients as such:
masterfs:/usr/local /usr/local nfs defaults 0 0

For some reason, one client's /usr/local is out of sync with the rest of clients (both files and the time stamps). For example:

On one client that is out of sync:

$ ls -all /usr/local/lcls/tools/edm/display/bpms
-rwxr-xr-x 1 softegr lcls 31991 Jul 25 2007 bpm_stripline_diagnostic.edl
-rw-rw-r-- 1 softegr lcls 32581 Apr 27 2007 bpm_stripline_diagnostic.edl~
-rw-rw-r-- 1 softegr lcls 6907 Dec 8 2006 bpm_stripline_sim.edl
-rw-r--r-- 1 softegr lcls 6943 Dec 8 2006 bpm_stripline_sim.edl~

On all others :

$ ls -all /usr/local/lcls/tools/edm/display/bpms
-rw-rw-r-- 1 softegr lcls 31991 Jan 16 16:27 bpm_stripline_diagnostic.edl
-rw-rw-r-- 1 softegr lcls 6907 Jan 16 16:27 bpm_stripline_sim.edl
-rw-rw-r-- 1 iocegr lcls 88455 Jan 30 09:49 bpm_values_table.edl
drwxrwxr-x 2 softegr lcls 4096 Jan 30 09:49 CVS


What is going on here? This is causing us a lot of problems. I would certainly expect that /usr/local should be transparent and in sync on all clients. But this dir on one client is clearly out of sync for quite long and doesn't know how to sync back with the rest automatically. I have to reboot the client to clear the problem when that happens.

Is there anyway to prevent this happen? Any help is greatly appreciated.

Thank you very much,
Jingchen

Possibly a product of having multiple clients mounted to the same file system editing the same file at the same time.

Are you using any version control software? You may have to use a file system with a locking feature or a version control system.