Frame-Relay Partial-Mesh question

[milkdroogy]

Hello!

my question is:

When you have frame-relay partial-mesh topology, something like this:

X

/ | \

/ | \

X X X

why everyone prefer to configure a subnet per vc(3 subnets in this case) instead of the simple solution of configuring 1 subnet and the top router as a multipoint?

is it a security issue or what?

[Edison Ortiz]

It's easier to configure routing protocols when creating subinterfaces.

With EIGRP and RIP, you have to deal with the split-horizon issue at the hub and with OSPF, you have to manually change the ospf network type at the interface or create neighbor statements under the OSPF process since the interface defaults to NON_BROADCAST.

I'm sure there more issues (multicast routing is another) but those mentioned are the first that came to my mind.

[purohit_810]

Partial mesh we are configuring for point-to-point and redundancy.

- DHarmesh