Chapter Summary

The IP protocol defines the details of routing and IP addressing. IP routing works similarly to the way that mail sorters work, and IP addresses have structure and meaning that are similar to the way that postal addresses work.

IP addresses are 32-bit numbers that are written in dotted-decimal notation. Each network interface needs an IP address. Any computer that has an IP address is called an IP host.

IP routing relies on the organization of IP addresses into networks or subnets. Without subnetting, all addresses in the same IP network reside on the same physical network. Routers can have a list of all IP networks, with instructions on how that router should forward IP packets to reach those networks. To do so, the router compares the destination IP address in the IP packet header with its list of network numbers.

With subnetting, all addresses in the same IP subnet reside on the same physical network. In effect, each subnet is treated as if it were a different IP network. So, routers can simply have a list of all IP subnets, with instructions on how that router should forward IP packets to reach those subnets. Also, by using subnets, fewer IP addresses are wasted.

IP networks can either be large (Class A), medium (Class B), or small (Class C). In a single Class A network, all addresses start with the same first octet, leaving 3 octets for the host part of the addresses. For Class B networks, all addresses start with the same first 2 octets, leaving 2 octets for the host part of the addresses. Finally, in Class C networks, all addressing starts with the same first 3 octets, leaving 1 octet for the host part of the addresses.