Frame Relay Technical Overview Necessity is the mother of invention, and this old rule applies to protocols as well. There were many shortcomings in X.25, SDLC, and other WAN protocols when Frame Relay surfaced. Frame Relay brought many useful options that aided in better network design, such as the following: -
Frame Relay provides a means for statistically multiplexing many logical circuits over a single physical circuit. -
Frame Relay helps lower line cost by not requiring dedicated end-to-end circuits for each link. -
Statistical multiplexing provides for increased network scalability by eliminating the need for a router serial port and CSU/DSU for each side of the connection. -
Frame Relay has a scalable network design: -
- It adheres to the three-layer model ” core , distribution, and access layers . -
- It allows full, partial, and hybrid meshing strategies. -
- It adds protocol broadcast and performance controls. Frame Relay is a CCITT and American National Standards Institute (ANSI) standard. It was founded as the next -generation protocol to X.25, which sometimes is referred to as an overengineered protocol because it performs extensive error checking and correction at the data link and network layers. X.25 had this type of error correction because it had to deal with many low-quality lines. Frame Relay implements a connection-oriented data stream that relies on the upper-layer protocols to provide error checking and correction. For reference, Frame Relay is outlined in the following standards: -
ANSI T1.606: "Architectural Framework and Service Description for Frame-Relaying Bearer Service" (1991) -
ANSI T1.617: "Signaling Specification for Frame Relay Bearer Service" (1991) -
ANSI T1.618: "Core Aspects of Frame Protocol for Use with Frame Relay Bearer Service" (1991) -
ITU Q.933 and Q.922: User plane control -
RFC 1490: Defines Frame Relay encapsulation A number of Frame Relay Forum (FRF) implementation agreements can be found at www.frforum.com. Frame Relay LMI Operation The LMI is essential for Frame Relay operation. When a Frame Relay link becomes active on a Cisco DTE device, it sends three LMI messages in rapid succession. The order is ANSI, ITU, and then Cisco. The router listens on DLCI 1023 for Cisco LMI and DLCI 0 for ANSI and ITU. The Frame switch responds with the LMI with which it is configured, and the router then sets the LMI type of that interface to match the LMI type that it received. If multiple types of LMI are received, the last one received is used. Cisco refers to this as LMI autosense. The router then sends LMI status messages back and forth every 10 seconds. These status messages are referred to as LMI keepalives . The router then begins to operate in the manner illustrated by Figure 5-3 and explained in the list that follows . Figure 5-3. Frame Relay LMI Operation  -
On every sixth LMI status request, the DTE device sends a full status request. This request serves as another keepalive and requests the frame switch to respond with a list of all DLCIs that have been defined for that link. Example 5-1 shows this initial exchange happening. Example 5-1 Initial LMI Setup and Exchange Output from debug frame lmi and debug frame packet 00:19:19: Serial0(out): StEnq, myseq 1, yourseen 0, DTE up 00:19:19: datagramstart = 0x400002DC, datagramsize = 14 00:19:19: FR encap = 0x00010308 00:19:19: 00 75 95 01 01 00 03 02 01 00 00:19:19: 00:19:19: Serial0(out): StEnq, myseq 1, yourseen 0, DTE up 00:19:19: datagramstart = 0x4000053C, datagramsize = 13 00:19:19: FR encap = 0x00010308 00:19:19: 00 75 51 01 00 53 02 01 00 00:19:19: 00:19:19: Serial0(out): StEnq, myseq 1, yourseen 0, DTE up 00:19:19: datagramstart = 0x400002DC, datagramsize = 13 00:19:19: FR encap = 0xFCF10309 00:19:19: 00 75 01 01 00 03 02 01 00 00:19:19: 00:19:19: Serial0(in): Status, myseq 1 00:19:19: RT IE 1, length 1, type 0 00:19:19: KA IE 3, length 2, yourseq 1 , myseq 1 00:19:19: PVC IE 0x7 , length 0x6 , dlci 110, status 0x0 , bw 0 00:19:29: Serial0(out): StEnq, myseq 2, yourseen 1, DTE up 00:19:29: datagramstart = 0x400002DC, datagramsize = 13 00:19:29: FR encap = 0xFCF10309 00:19:29: 00 75 01 01 01 03 02 02 01 00:19:29: 00:19:29: Serial0(in): Status, myseq 2 00:19:29: RT IE 1, length 1, type 0 00:19:29: KA IE 3, length 2, yourseq 2 , myseq 2 00:19:29: PVC IE 0x7 , length 0x6 , dlci 110, status 0x0 , bw 0 -
When the frame switch receives a status inquiry message, it sends a full status response. This message sends a list of all active DLCIs on that port. -
For each active DLCI, the router sends an Inverse ARP request per Layer 3 protocol configured on that interface. If IP and IPX are configured, for example, the router sends two Inverse ARP requests. The request also asks the other router to reply with its network layer address. If Inverse ARP is not supported, this transaction and the transaction that follows are accomplished with the frame-relay map commands (covered in the "Configuring Frame Relay" section). -
For each DLCI that each router receives an Inverse ARP message about, the router creates a map entry in its Frame Relay map table. The map table includes the local DLCI and the remote router's network layer address for that request. The table also contains the status of the PVC, which is one of the following (as displayed with the show frame-relay pvc command): - - ACTIVE ” Indicates that the PVC is active and that information can be exchanged
- - INACTIVE ” Indicates that the local connection to the Frame switch is working, but the remote router's connection to the frame switch is not working
- - DELETED ” Indicates that no LMI is being received from the frame switch or that the physical layer is still not established
-
The router continues to exchange keepalive messages every 10 seconds. Again, every 60 seconds, or sixth exchange, a full status LMI request is sent and the process repeats itself. If three consecutive LMI messages are missed, the link is brought down. |
