The main point is that the standard is that a phsycial network segement
can have only one IP network on it. If there are more than one network,
then one of the machines must act as a router and route traffic from one
network to the other and handle the split horizon issues for RIP. The end
result is increased traffic on the network segment and increased complexity
and processing requirements for the hosts and routers.
Something to consider is that some host implementations of secondary
addressing is incomplete. I've seen several cases where a portmaster
sends a packet to a host at one IP address on the same network and then
recieve a reply from a DIFFERENT IP address on a DIFFERENT network. The
host didn't bother to set the source address to match the network of the
destination address.
Larger ISPs use dynamic Ip address assigned as pools to PMs. The assigned
address pools can be in a different network (or subnetwork) from the ethernet
segement because the PMs become routers to those addresses.
You can ofer static IPs over multiple PMs. What you have to do is set the
netmask of the IP address to 255.255.255.255. This way the PM will send
the RIP update with the static IP as a host route. Obviously this will
greatly increase the RIP traffic over the network and create larger
routing tables.
JGT
-- John G. Thompson Livingston Enterprises Inc. Phone: (800) 458-9966 JOAT(MON) 6920-220 Koll Centre Pkwy. Fax: (510) 426-8951 support@livingston.com Pleasanton, CA 94566 http://www.livingston.com