frame relay subinterfaces using point to point vs subinterfaces vs phsyical interface

orange_badger

if someone would be so kind to please explain to me when and why a frame relay config would need to use these interface types.

im going through my cisco press icnd 2 book and it tells me that on a full mesh topology with 3 routers each having two pvc's the most basic setup would be .

encapsulation frame relay (interface subcommand )

ip address (interface sub command)

i thought if a router had more then 1 pvc it would need to use a sub interfaces so it could have a dlci for each pvc

thanks

JeanM

My understanding with frame relay full mesh, all routers would have virtual circuits to all destinations. Total number of links would be n(n-1) / 2 where n is the number of nodes. Then you also have PVC (perm virtual circuit) and in partial mesh where you only use one subnet per pvc you would use 1 single physical interface on the router , for example s0/1 but have it configured for sub-interfaces and each one with different dlci for point to point, such as s0/0/0.201 would for dlci of 201, and s0/0/0.203 for dlci of 203 to the other frame relay routers...

Zartanasaurus

The main interface is already a multipoint interface. There's no requirement to have subinterfaces for multiple VCs. Use subinterfaces if you need different subnets.