Hello...
I am looking for some help with the following scenario and would appreciate
your input.

Environment:

Master Server: NT 4.0 SP 6a. Netbackup patch 363.
Media/Slave: NT 4.0 SP6a NetBackup patch 363.
Client: HP-UX 11.0

Problem:

I get a Error 54 (timed out connecting to client) when I try to perform
backup of the Unix Client to the NT media server.

The bpcd log on the client shows:
22:47:18 [1963] <2> bpcd main: offset to GMT 28800
22:47:18 [1963] <2> bpcd main: setup_sockopts complete
22:47:18 [1963] <2> bpcd main: output socket port number = 536
22:47:18 [1963] <8> bpcd peer_hostname: gethostbyaddr failed :
HOST_NOT_FOUND (1)
22:47:18 [1963] <16> bpcd peer_hostname: gethostbyaddr failed to return
peer host, herrno = 1
22:47:18 [1963] <16> bpcd main: Couldn't get peer hostname

The media server has 4 network cards and one unique hostname. When I ping
the media server by hostname from another machine, I consistently get
response from the same ip address - not from any of the other three.  I get
the same ip address also, when I issue a bpclntcmd -hn ninja (media server)
from the master server. This leads me to believe that the media server
primarily responds on the same ip addr. This brings two questions to mind:

1. What entry do I put for the REQUIRED_INTERFACE parameter on the media
server so that it will always respond on the same interface? Is it the ip
address that I want it to respond on?
2. How do I make sure that the hostname ninja will resolve to the ip
address of the REQUIRED_INTERFACE consistently when a client attempts to
contact the media server?

Appreciate all your help!

Vik
.....................................................................................................

Vik Bhatia
Database Administrator
Sony Computer Entertainment America
919 E Hillsdale Blvd. Foster City, CA 94404
650-655-8069
.....................................................................................................