We have a problem with certain jobs being launched automatically by bpsched.  
The jobs that hang or are delayed use a storage unit that is not on the master 
server.  When a successful job is kicked
off it looks like the master server communicates to bpcd on the media server 
and bpcd then contacts the client and launches the backup job.

The jobs that  are delayed or fail entirely operate differently.  bpsched finds 
a job scheduled to run, and instead of contacting bpcd on the media server it 
attempts to contact the client directly.
The master server does not have a network connection to the client, so bpsched 
cannot contact the client.  bpsched will then try a different port address to 
contact the client and repeats this process
over and over again.  Sometimes it magically stops and contacts bpcd on the 
media server and backs up the clients, other times it just hangs.   Any 
suggestions.

Selwyn Schultz
Herman Miller Inc.
e-mail:  Selwyn_Schultz AT hermanmiller DOT com
Phone:  616 654-7257
Fax: 616 654-5295