OSPF Areas

One of the key reasons for the rapidity of OSPF's convergence is its use of areas. Remember that the two main goals that the IETF sought to achieve with OSPF were

• Improved network scalability

• Rapid convergence times

The key to both goals lies in compartmentalizing a network into smaller regions. These regions are known as areas. An area is a collection of networked end systems, routers, and transmission facilities. Each area is defined with a unique area number configured into each router. Router interfaces defined with the same area number become part of the same area. Ideally, these areas are not arbitrarily defined. Instead, the boundaries of an area should be selected so as to minimize the amount of traffic between different areas. In other words, each area should reflect actual traffic patterns rather than geographic or political boundaries. Of course, this is a theoretical ideal and may prove impractical in your particular environment.

The number of areas an OSPF network can support is limited by the size of its Area ID field. This field is a 32-bit binary number. Therefore, the theoretical maximum number of networks is a 32-bit binary number with all its bits equal to 1. The decimal equivalent of this number is 4,294,967,295. Obviously, the practical maximum number of areas you can support is much less than this theoretical maximum. In practice, how well designed the network is will determine the practical maximum number of areas you can support within it. Figure 12-1 illustrates a fairly simple OSPF network with just three areas, numbered 0, 1, and 2.

Figure 12-1: A small OSPF network with three areas.

0 0

Post a comment