160Appendix E: Understanding IP and IP addressing
representative to discuss
Variable-width subnetworks
When subnetworks were first invented, they were intended to be used in a star topology, with the major router at the port of entry connected to all subnetworks. All subnetworks were supposed to have address ranges of the same size. Later IP implementations have retained the expectation that the width of the subnet mask is uniform throughout all the subranges of the top level (class A or B) network number.
With the current increased emphasis on conservation of IP addresses, it is often desirable to allocate subnet ranges of “just the right size”. To allocate ranges consistently, all protocol exchanges that communicate a network address range must include the associated subnet mask. OSPF performs this, and future releases of other route information protocols (for example, RIP Version 2) allow this too. On the other hand, some protocols do not carry this information, since knowledge of subnet structure is contained within a routing domain, and is invisible outside of a routing domain.
It is important to realize that support for
Protocols related to Internet Protocol
This section provides brief descriptions on the constituent protocols of TCP/IP.
Internet Protocol
IP is a connectionless datagram service that provides the following benefits:
•
•
•fragmentation and reassembly
•
•checksum verification of header contents
IP is defined in RFC791.
