In a large, enterprise-level internetwork, choosing the correct routing protocols is imperative. Network changes and growth greatly affect the best path for routing packets through an internetwork.
Routing protocols are responsible for keeping track of changes on the network and for sharing these changes with routers throughout the network. Routing protocols are categorized into two major types: distance vector and link state.
Distance vector routing protocols are designed to run on small networks (usually fewer than 100 routers). Examples of distance vector routing protocols include RIP and IGRP. Distance vector protocols are generally easier to configure and require less maintenance than link state protocols.
On the downside, distance vector protocols do not scale well because they require higher CPU and bandwidth utilization. They also take longer to converge than do link state protocols.
Distance vector routing protocols use a hop count to determine the best path through an internetwork. The hop count is simply a measure of the number of routers a packet must cross to get from host A to host B. For example, if host A is attempting to telnet to host B, and the packet must cross three routers to get to host B, then the hop count is three.
Distance vector protocols always choose the route with the fewest number of hops as the best route. This can be a problem when the best route to a destination is not the route with the least number of hops. For example, suppose host A is trying to connect to host B and there are two paths available: Router A is using a T-3 connection and Router B is using a dial-up 28.8 Kbps connection. If Router B is one hop away, but Router A is two hops away, Router B will be chosen as the best route, even though it’s not the fastest alternative.
Convergence time is the amount of time it takes to propagate changes in network topology throughout the internetwork. Distance vector routing protocols do this by periodically sending out the entire routing table to every router on the network. In a large environment with many routers, this can greatly affect CPU and bandwidth utilization.
Distance vector routing protocols are great for a small environment, but when it comes to enterprise networking, you must deploy link state protocols.
Link state routing protocols are designed to operate in large, enterprise-level networks. Examples of link state protocols include OSPF and NLSP. Link state routing protocols are very complex and are much more difficult to configure, maintain, and troubleshoot than distance vector routing protocols. However, link state routing protocols overcome many of the shortcomings of distance vector protocols.
Link state protocols use a different algorithm than distance vector protocols for calculating the best path to a destination. This algorithm takes into account bandwidth as well as other factors when calculating the best path for a packet to traverse the network. Additionally, link state convergence occurs faster than distance vector convergence. This is because link state establishes a neighbor relationship with directly connected peers and shares routing information with its neighbors only when there are changes in the network topology.
This method of exchanging routing information is different from that used by distance vector routing protocols, which periodically (about every 90 seconds) send their updates to every router on the network. Also, link state routing protocols only send updates to neighboring routers, unlike distance vector protocols, which send the entire routing table. Link state routing protocols can be difficult to configure and maintain, but the efficiency your network will gain through their use is worth the time and effort.
Choosing the routing protocol that’s best for your network can be a daunting task. For more information on choosing and configuring routing protocols, check out CCIE Professional Development: Routing TCP/IP, Volume I .
If you'd like to share your opinion, please post a comment below or send the editor an e-mail .
Warren Heaton, CCDA, CCNA, MCSE+I, is the Cisco Program Manager for A Technological Advantage in Louisville, KY.