Hi Adrian,
Using two RIS servers on the same network means whichever RIS server the
DHCP server gives to the client is the server the client will use. In the
event that the DHCP server does not give a PXE boot server IP then the
client will broadcast for one; this means you can't select the RIS server at
the client level.
You can prestage your workstations on the RIS server so they go to the
correct RIS server (preventing RIS'ing across a WAN for example).
My experience has shown that in a WAN environment, you want to put RIS on
the DHCP server in the remote subnet so that the DHCP offer has a PXE boot
server IP that is its own IP, this helps eliminate the possibility that the
client will try and RIS across a WAN. You can also control the DHCP server
by the helper address that you use in your routers - point them to the
correct server.
I wish MS would have included a site-aware RIS server in 2003 like they did
with the greatly-improved DFS in W2003.
--
Scott Baldridge
Windows Server MVP, MCSE
"Adrian Koch"
