Frame-Relay explained

The other night i showed some frame-relay sample configuration and today i am going to blog more about frame-relay networks. Frame-relay is a reliable and inexpensive WAN protocol whereby it provides QoS service using DE bit, FECN and BECN bit. It also save cost for company who used to deploy leased-lines in the past by creating Permanent Virtual Circuit(PVC) with physical access lines. Before going further, below are some terms used the in Frame-relay terminology. I will not write a full article on frame-relay because the whole concept is so huge. I will just briefly describe how it works and its usage.

CIR Committed Information Rate [Guaranteed rate at which the network commits to transfer user data under normal conditions]
EIR Excess Information Rate [Maximum rate capactiy on top of CIR]
DLCI Data Link Connection Identifier [PVC end point connection identifier with only local significance]
LMI Local Management Interface [Maintenance protocol for frame-relay]

How frame-relay works is when the router forwards data to the frame-relay switch, the switch in turn forward the frames to the correct destination over a permanent virtual circuit. At each end of the PVC is identified with a DLCI value with only local significance value. A DLCI value is a 10-bit address in the frame-relay header which provides point-to-point or point-to-multipoint connection. I had drawn a simple diagram which depicts the explaination. For site A to send site B packets, it simply specifies the appropriate DLCI number for the virtual circuit that connects to Site B site in the Frame Relay header. However, please note that the DLCI value does not identify the whole PVC network, instead it is just the connection from your router to the frame-relay switch and when it reaches its destination, the DLCI might be a different value. That explains why it has only local significance. The CIR is the traffic rate at which it is guaranteed by your ISP. If packets are sent faster than the cIR rate which you had agreed with your ISP, your ISP might drop all or some of the excess packets depending on how the network is setup. A DE bit is set in the frame-relay header to indicate excess flow of the packets. So if the network is not congested, the packets will flow through, else the excess packets will be discarded. FECN and BECN are flags set in the frame-relay header to allow control of the congested traffic. If congestion is encountered during the traffic flow, the FECN flag will be set on the switch to indicate that there will be a delay of packets arrival and vice versa with the BECN flag set on the receiving frame-relay switch. In this manner, both the sending and the receiving router can expect congestion in the carrier network and delay of packets upon arrival. The LMI protocol is used for keepalive purposes and global addressing purpose like making the DLCI value to have a global significance. As you can see, i am just explaining in a very very basic manner, because the whole technology is so huge that you can even publish a entire book for it. Configuration wise, i had already shown you guys a few examples with basic and simple frame-relay setups. You guys can read more if you are interested. Just search in wiki or cisco.com and a wealth of information is waiting for you to explore. However, i am still exploring more options in terms of the configuration area.

How about spoofing frame-relay frames and reroute the whole traffic to a bogus network? Personally i had not tried it before because of cost and lack of resources. Secondly, i guess i do not touch frame-relay networks often. Well, if you guys knows how to do frame-relay rerouting or spoofing, let me know.