Administering Cisco QoS for IP Networks

Tweak and tune your Cisco routers for best performance with this practical guide.
Visit www.syngress.com/solutions to have your questions about this chapter answered by the author.
| Q. | Is EIGRP as simple to implement as IGRP? | |
| A. | No. I have read that EIGRP is IGRP with an E . This is absolutely not true. The underlying mechanisms of EIGRP are vastly different and must be considered when designing an EIGRP network. As stated earlier, the only similarity between the two protocols is their use of the composite metric. If EIGRP is implemented in the same manner in which IGRP would be used on the same network, many of the advantages of EIGRP are tossed out. | |
| Q. | Why is a diffused computation better than simple routing by rumor? Seems like a lot of extra work for the same result. | |
| A. | The result is not the same. While IGRP and EIGRP may select the same path, the guarantee of a loop-free topology created by a diffused computation outweighs the burden of understanding the complexity of the protocol. An IGRP process can never guarantee a loop-free topology; it can only attempt to prevent loops. | |
| Q. | If such things as stuck-in-active (SIA) routes are so difficult to troubleshoot, why not use another protocol? | |
| A. | In a well-designed and well-thought-out network, SIA events should not occur. If hierarchical summary addressing is implemented, query boundaries can be created that limit the propagation of diffused computations to only the routers that could possibly know another route. Congested, slow WAN links that are used to... |