we have a mixed cluster environment with NW 6.5 and OES2 Linux servers.
There are mostly in the morning access problems with our DFS volumes (EMC).
The Windows XP SP2 clients with NW CL 4.91 SP4 have the error, that no access is possible, and that the harddisk is not formated. Also the login scripts does not map all drives. Our ideas of making a VLDB repair are not the solution.
There were 2 replicas of the VLDB on 2 OES2 Linux Servers. So we decided to fix the problem, by making only one replica. But after some minutes we have the same problems and the same messages on the clients. Server Reboot brings success for some minutes and then access is aborted furthermore.
The situation is frustrating because you can only see some errors in /var/log/messages in jstcpd. And there are now TIDs and now entries in forum.

Has anybody some ideas?