Frame Relay physical interface vs. multipoint interface

jscimeca715

Let's say that RouterA has a multipoint connection and one point-to-point connection to a frame relay network. Is it okay to configure the multipoint connection using the physical interface, bypassing the need to create a multi-point subinterface? I could then create a subinterface under that physical one for my point-to-point VC? Or, if you have a multipoint connection coupled with a point-to-point connection do you need to use the subinterfaces only and leave the physical interface alone?

Thanks for the help!

Neeko

jscimeca715 wrote: »

Let's say that RouterA has a multipoint connection and one point-to-point connection to a frame relay network. Is it okay to configure the multipoint connection using the physical interface, bypassing the need to create a multi-point subinterface? I could then create a subinterface under that physical one for my point-to-point VC? Or, if you have a multipoint connection coupled with a point-to-point connection do you need to use the subinterfaces only and leave the physical interface alone?

Thanks for the help!

Correct, the physical interface has to be unassigned if any subinterfaces are to be configured. The physical interface does require the encapsulation frame-relay command however.

jscimeca715

Excellent, thank you Neeko. I have one more question that is preventing me from really understanding Frame Relay fully. When do you need to use the frame-relay interface-dlci dlci-# command? I understand you need it when configuring point-to-point subinterfaces. Does it also apply to multi-point as well? The only reason I ask is because in one example in Odom's book all he does is configure the encapsulation frame-relay command on the physical interface for all three routers in a full mesh topology. That's with all defaults assumed. So my understanding is this: In the full-mesh topology he was relying on the frame-relay switch to generate LMI messages stating which PVC is up. Does the Frame-Relay switch also do that for subinterfaces? In production networks? Any clarity will be appreciated.

ncsugrad2002

jscimeca715 wrote: »

Excellent, thank you Neeko. I have one more question that is preventing me from really understanding Frame Relay fully. When do you need to use the frame-relay interface-dlci dlci-# command? I understand you need it when configuring point-to-point subinterfaces. Does it also apply to multi-point as well? The only reason I ask is because in one example in Odom's book all he does is configure the encapsulation frame-relay command on the physical interface for all three routers in a full mesh topology. That's with all defaults assumed. So my understanding is this: In the full-mesh topology he was relying on the frame-relay switch to generate LMI messages stating which PVC is up. Does the Frame-Relay switch also do that for subinterfaces? In production networks? Any clarity will be appreciated.

You def need it on both. I would strongly recommend watching the CBT Nugget video on this subject. He goes over in depth advantages/disadvantages of each type of frame relay setup and a configuration of each in everything from single links to each site in a hub and spoke style setup to a fully meshed environment where everyone connects to everyone. It's really a great video--I didn't understand it until after I watched it.

networker050184

Neeko wrote: »

Correct, the physical interface has to be unassigned if any subinterfaces are to be configured. The physical interface does require the encapsulation frame-relay command however.

That is not correct. You can use the main interface for a multipoint connection and a sub interface for a point to point connection at the same time.


To the OPs other question you do not need the frame-relay interface-dlci dlci-# on multipoint sub interfaces if you are using inverse arp.