Frame Relay Theory

I finished going through the Frame Relay chapter in the ICND2 which helped me further understand this WAN technology. This is probably the only other chapter besides subnetting that tripped me up a little bit. More specifically it has to do with the fact that Frame Relay needs to use both Layer 2 and 3 protocols. At the layer 2 it took me a while to get the DLCI global and local theories nailed down. Also knowing when one protocol/technology begins and ends in a Frame Relay topology. For example LMI (Local Management Interface) provides keep alives between the access links.

But encapsulation occurs between the routers on the PVC (Permanent Virtual Circuit) between each other, so it crosses the Frame Relay network. Where as LMI's only cross between the DTE and the DCE which is usually the Router and Frame Relay switch at the ISP. At layer 3 you have determine how you want to configure the interface IP addresses. You can actually configure the Frame Relay to use one subnet if it's in a full mesh. However if it's only in a partial mesh it's probably best to use seperate subnets for each PVC, this is especially true if you're using a distance-vector routing protocol. You can also configure a hybrid of the two options. All in all I'm going to spend a little extra time with Frame-Relay to get it down by exam time.