Frame Relay Applications: TCP/IP Suite - Page 4
Inverse ARP
Inverse ARP will be discussed here as it applies to IP networking and address discovery. Inverse ARP operates in a similar fashion for Frame Relay DLCI discovery for AppleTalk, Banyan VINES, DECnet, Novell IPX, and Xerox Network Services (XNS).
The motivation for the development of Inverse ARP is a result of the desire to make dynamic address resolution within Frame Relay both possible and efficient. PVCs and, eventually, SVCs are identified by a DLCI. These DLCIs define a single virtual connection through the WAN and are the Frame Relay equivalent to a hardware address.
Periodically, through the exchange of signaling messages, a network might announce a new VC with its corresponding DLCI. Unfortunately, protocol addressing is not included in the announcement. The station receiving such an indication will learn of the new connection, but will not be able to address the other side. Without a new configuration or mechanism for discovering the protocol address of the other side, this new VC is unusable. RFC 1293 defines Inverse ARP in more detail.
Whereas ARP enables a system to build a table mapping the LAN system's IP address to the Layer 2 MAC address, Inverse ARP is used to build a similar table mapping the connected system's IP address (by Frame Relay Virtual Circuit) to the Layer 2 DLCI on the connected system.
Figure 15-20 illustrates a four-node Frame Relay WAN, each site interconnected by a Frame Relay PVC.
Figure 15-20: Four-Node Frame Relay WAN with DLCI Assignments
(Click image for larger view in a new window)
Router B needs to determine the IP address of its neighbors prior to the forwarding of traffic across the interconnection. Router B also needs to map each neighbor's IP address to the DLCI that will be used to reach that neighbor. Router B essentially needs a table similar to Table 15-10.
Table 15-10: Router B's Inverse ARP Table
| IP Address | DLCI |
| 172.16.24.102 | 37 |
| 172.16.24.100 | 25 |
| 194.213.14.214 | 17 |
The IP Address column identifies the IP address of Router B's neighbor; the associated right column identifies the corresponding DLCI assignment.
Although this table could be built manually, it is far more efficient to let the Inverse ARP mechanism build it. Each router uses a simple two-step procedure to build this table. These steps include the following:
- Each router sends a message across each connected VC asking for the IP address of the distant end.
- The sending router then records the IP address and the corresponding DLCI into its tables.
NOTE: ARP is used when the Layer 3 address is known but the corresponding Layer 2 address is unknown, typically the MAC address. Inverse ARP is used when the Layer 2 address is known, typically the DLCI, but the corresponding Layer 3 address is unknown.
