Summary

One way of classifying routing protocols is by the type of algorithm they use to determine the best path to a destination network. Routing protocols can be classified as distance vector, link state, or path vector. Distance vector means that routes are advertised as vectors of distance and direction. Distance is defined in terms of a metric such as hop count and direction is simply the next-hop router or exit interface.

Distance vector routing protocols include:
RIPv1
RIPv2
IGRP
EIGRP

Routers that use distance vector routing protocols determine best path to remote networks based on the information they learn from their neighbors. If Router X learns of two paths to the same network, one through Router Y at 7 hops, and another route through Router Z at 10 hops, the router will choose the shorter path using Router Y as the next-hop router. Router X has no knowledge of what the network looks like beyond Routers Y and Z, and can only make its best path decision based on the information sent to it by these two routers. Distance vector routing protocols do not have a map of the topology as do link state routing protocols.

Network discovery is an important process of any routing protocol. Some distance vector routing protocols such as RIP go through a step-by-step process of learning and sharing routing information with their neighbors. As routes are learned from one neighbor, that information is passed on to other neighbors with an increase in the routing metric.

Routing protocols also need to maintain their routing tables to keep them current and accurate. RIP exchanges routing table information with its neighbors every 30 seconds. EIGRP, another distance vector routing protocol, does not send these periodic updates and only sends a "bounded" update when there is a change in the topology and only to those routers that need that information. EIGRP is discussed in a later chapter.

RIP also uses timers to determine when a neighboring router is no longer available, or when some of the routers may not have current routing information. This is typically because the network has not yet converged due to a recent change in the topology. Distance vector routing protocols also use triggered updates to help speed up convergence time.

One disadvantage of distance vector routing protocols is the potential for routing loops. Routing loops can occur when the network is in an unconverged state. Distance vector routing protocols use holddown timers to prevent the router from using another route to a recently down network until all of the routers have had enough time to learn about this change in the topology.

Split horizon and split horizon with poison reverse are also used by routers to help prevent routing loops. The split horizon rule states that a router should never advertise a route through the interface from which it learned that route. Split horizon with poison reverse means that it is better to explicitly state that this router does not have a route to this network by poisoning the route with a metric stating that the route is unreachable.

Distance vector routing protocols are sometime referred to as "routing by rumor", although this can be somewhat of a misnomer. Distance vector routing protocols are very popular with many network administrators as they are typically easily understood and simple to implement. This does not necessarily mean link-state routing protocols are any more complicated or difficult to configure. Unfortunately, link-state routing protocols have received this somewhat unwarranted reputation. We will learn in later chapters that link-state routing protocols are as easy to understand and configure as distance vector routing protocols.