Frame Relay, InARP

knownasthatguy

Ok, my cisco class was having a discussion on FR and this topic was brought up by me:

If InARP is used to map known MACs to DLCI's, then how does the DTE already know the MAC of the attached FR switch. Here is an excerpt of the InARP RFC:
When an interface supporting InARP becomes active, it should initiate
the InARP protocol and format InARP requests for each active PVC for
which InARP is active. To do this, a requesting station simply
formats a request by inserting its source hardware, source protocol
addresses and the known target hardware address

Now, my question is the known target hardware address. How does or when does the DTE get this information?

knownasthatguy

From what i can gather, when the pvc is first built into the dte and the dte connects to the fr switch, the initial LMI status message sent to the swtich will retrieve the appropriate information, like the hardware address. since fr is a non broadcast, arp and rarp can not be used, so InARP was developed to overcome this limitation. InARP only acts upon what it knows in hardware addresses while LMI retrieves the HW addy prior to InARP. I guess one could argue that there are two layer two protocols at work, the LMI which goes first and then the InARP, if needed, goes next. It could be thought of as the LMI is only concerned with L2 so it would only be logical that it goes first, and while InARP goes from L2 to L3, it is a dependent variable to the LMI and would have to wait until LMI fisnishes its L2 function.

Of course, this was ascertained through best guess assumptions and intuition from reading RFCs, text books and discussions. If anyone has some info, hollar back