Illustrated TCP/IP by Matthew G. Naugle Wiley Computer Publishing, John Wiley & Sons, Inc. ISBN: 0471196568   Pub Date: 11/01/98

Chapter 84
Classless Inter-Domain Routing (continued)

You know that I must be leading up to something. It is the next step in understanding IP addresses and Internet routing. It is called CIDR (pronounced cider). CIDR is explained in RFCs 1517–1520, so I am not detailing the CIDR spec here. Just the concept. The concept is simple: Implement a generalization of Variable Length Subnet Masks and move from the traditional Class A, B, C address toward the idea of a 32-bit IP address and a prefix (without the concept of a Class). In CIDR, there are 32 bits and a prefix. To understand CIDR, you must place the concept not on your local network but on the Internet routers. You can employ CIDR on your network, but there is really no reason to (since your hosts would have to be configured to understand supernets). The Internet routing tables were expanding at a exponential rate (without CIDR, they would have passed over 80,000 routes today). The Internet routers are simply those devices that move data towards a destination indicated by its IP address, and therefore do not have large subnets off of them with which to support hosts. CIDR works on the notion that we are routing arbitrarily sized (a range) network address space instead of routing on Class A, B, and C. CIDR routes based on routing information that has the prefix attached to it. For example, the address of 200.15.0.0/16 could be an entry in the Internet routing table—one entry indicating a range of addresses. Any IP datagrams received by that router with the first 16 bits indicating 200.15 would be forwarded out the port indicated in the routing table. This prefix could be assigned to any range of addresses because CIDR does not associate a prefix with a Class.