OSPF (Open Shortest Path First) Protocol

OSPF (Open Shortest Path First) Protocol

The Open Shortest Path First (OSPF) protocol is the most widely used example of a link state routing
protocol. It is in wide use as an interior gateway protocol on the Internet and many other networks. The latest version of this public protocol was version 3, as specified in RFC 5340 released in 2008, and includes support for IPv6.

The Open Shortest Path First algorithm operates similarly to Dijkstra’s algorithm but adds a system of designated (primary) and backup routers. Routers are selected for these roles based on their priority number; routers with a priority of 0 cannot be designated or backup routers. The designated router for an area is responsible for sending Link State Advertisements (LSAs) to all other area nodes. OSPF routing packets on an OSPF routed network have a nine-field header, illustrated in
Figure 9.9. OSPF packet types include HELLO, database description, link state request, link state update, or link state acknowledgment.

OSPF is used on autonomous systems (AS). Autonomous systems are one or more networks under a common administrative structure. OSPF functions not only as the interior gateway routing protocol for the AS, but it can also send and receive routes from other autonomous systems. Each network in the AS is an area within a hierarchy defined within the AS, each area being a collection of contiguous hosts. In OSPF, a routing domain is an alternative description for all systems in an AS
that share the same topological map. OSPF partitions areas into separate topologies so that each area is kept unaware of another area’s routing traffic. This system is meant to lower the amount of overall network traffic and speed up the discovery process of shortest routes for an individual area.

Collections of areas are connected by OSPF border routers in an OSPF backbone. The backbone itself is organized as an OSPF area, and routing information for that area is also separate from the areas the backbone connects. It is possible to organize an OSPF backbone so that the backbone is composed of two or more unconnected groups. The backbone is made contiguous by defining a virtual link through routers in a non-backbone area to serve as the connection between backbone groups. The backbone of an OSPF system composed of border routers communicates with other exterior gateway protocols (EGPs) such as the Border Gateway Protocol (BGP) or the Exterior Gateway Protocol (EGP). Figure 9.10 shows an OSPF network with several areas, a backbone, and a virtual link.

Share this post

Post Comment