standards

This is the best book I've read on IGP Routing Protocols and their implementation on Cisco Routers. The book takes a reader from basic IP routing fundamentals through to in-depth implementation of the most complex IGP routing protocols. The book is easy to read and flows in very logical manner. I especially enjoyed the chapter (8) on Cisco's EIGRP; It was the most detailed coverage I have read on Cisco's proprietary Interior routing protocol.

Additional Comments:

I did note the following errors in the books print listed by chapter:

Chapter 5

Page 193, Paragraph 1, Sentence 4 and 5

The statement with regard to holddown timers and Routers running RIP routers is false. The statement read "If an advertised hop count is higher than the recorded hop count and the update was received was originated by the recorded next-hop router, the route will be marked as unreachable for a specified holddown period. If at the end of that time the same neighbor is still advertising the higher hop count, the new metric will be accepted". The reason this is wrong is because, a router that receives a route with a higher metric than the currently recorded metric and the advertising router is also the recorded next-hop router, the router will immediately install the new metric. (lab tested).

Appendix F

Page 970, Chapter 7, Answer 1

The given answer is not completely correct as stated. First the answer states that the router labeled C on page 321 and who's configuration is on page 324 will interpret the RIPv2 routes 192.168.90.0/29 and 192.168.13.86/28 coming from routers A and B respectfully as 192.168.13.64/27 and discards them because this route is directly connected to one of router C's interfaces. This is incorrect first because the routes that will be advertised by routers A and B will be 192.168.13.80/29 and 192.168.13.80/28 respectfully not 192.168.13.86/28 and 192.168.90.0/29. Secondly because router C is configured with the default rip settings, which allows the router to advertise RIPv1 and receive RIPv1 and RIPv2, which will allow router C to correctly interpret the masks sent with the RIPv2 updates sent by routers A and B as such router C will have both (Lab Tested).

Appendix C, Chapter 11, Answer 1

The answer as stated is not completely correct. The answer states that in order to redistribute between the IGP domain and the RIPv1 domain the router labeled B should have its mask changed from /27 to /28. Although this facilitates the routes from the IGRP domain into the RIPv1 domain it does not completely facilitate the reverse. The reason for this is that when router B receives the RIPv1 updates from router C it interprets those routes to have a /28 mask due to the new mask on its E1 interface. This results in the IGRP domain believing that the subnets in the RIPv1 domain have only 14 hosts per subnet where as they really can have as may as 30 hosts. In example take the subnet 172.16.2.32/27 which possibly contains hosts 172.16.2.33 -62 in the RIP domain, when this route is redistributed into the IGRP domain it would be interpreted to only contain hosts 172.16.2.33-46. This would cause packets originating in the IGRP domain destined for the hosts numbered 172.16.2.47-62 to be returned unreachable by the routers in the IGRP domain without further configuration. i.e. static routes.

I've just finished reading Jeff Doyle's "Routing TCP/IP Volume I" book and feel confident that I can now pass the CCIE lab in the near future. His presentation of TCP/IP IGP routing protocols is very clear and concise, with a perspective and insight into internetworking which is rarely encountered. This edition combines introductory, historical, conceptual and practical materials, and RFC's, with a first rate writing ability into one scintillating package.

The chapter on OSPF was particularly satisfying going into the depth where the next step would be research into the protocol. Having said that there is a very pragmatic side to this book including configuration examples within the text as well as configuration and troubleshooting exercises and review questions at the end of the relevant chapters.

By leaving nothing on IGP's out and all the while remaining emminently readable, this book marks a very welcome and refreshing addition to the internetworking literature and will no doubt become known as a seminal work. I truly appreciate what I have learnt from this book, and am very much looking forward to the publication of Jeff's next work on EGP's, in Volume II, in July.

I am reviewing Routing TCP/IP Volume 1 (ISBN 1578700418), part of the CCIE Professional Development Series from Cisco Press. This book is widely regarded as part of the "Holy Trinity" of CCIE preparatory books. In addition to CCIE prep, it is also highly valued as an essential desk reference for anyone pursuing a career as a senior-level routing engineer. The author, Jeff Doyle, having written the two most recognized tomes on IP Routing, is a respected authority on the topic. His writing style is very clear and not at all difficult to read, which sets him apart from a lot of the authors in the "high-end routing book" category.

The book is part of a set of two books. TCP/IP Volume 1 contains coverage of the major interior routing protocols (RIP, IGRP, EIGRP, OSPF & ISIS) and follows it with coverage of route redistribution, filtering and mapping. TCP/IP Volume 2 covers the BGP exterior gateway protocol and follows it with coverage of advanced IP routing issues. As someone who's come up through the CCNA-to-CCNP-to-CCIE-candidate path, I recognized all of the topics from the BSCI curriculum. All this book really does is takes them to another level of depth. I highly recommend this path of coming at the CCIE, because unless you've done it all and seen it all, there's way too much information to take in during a single reading. Reading the BSCI book first gets you familiar with all of the topics, so that you're not overwhelmed when reading the Doyle books.

While many pan this book as being outdated since it was written in 1998, my contention is that all of the covered topics are still fair game for the CCIE qualifying written exam and the book still retains all of its original value. There are a lot of topics which are on the test which are not in Doyle's books, but if you look at the CCIE blueprint, the topics covered in the books map exactly to the topics in the IP Routing & IP Multicast sections of the blueprint. You can't treat any book as being a one stop shop for CCIE preparation. That's impossible. The book would be the size of my desk. I would actually contend that, on the next rewrite, they should break the 2 volume set into 3 volumes, one for the interior protocols, one for BGP by itself, and one for all of the advanced routing issues and multicast. This would make the volumes a bit less daunting and also somewhat easier to carry.

But, carry-ability issues aside, this is a very good book. It takes all of the topics to the degree of depth which you would expect for a CCIE-level book and explains things in a way that doesn't lose the reader. I had to work to keep the examples straight in my head, but no one said becoming a CCIE was easy. On the downside, the book contains no disc. This is unfortunate, because I always like to have the book on PDF. Also, it would be nice to have the review questions and answers put into a question bank. The book is also missing the "Do I Know This Already" section for each chapter, which is present in all of Cisco Press' certification guides. I realize that this is probably due to the fact that this is more of a desk reference / learning book than an official certification guide. However, as someone who's studying for an exam, I always like as many free practice questions as I can get.

I give Routing TCP/IP Volume 1 a 5 on my 5 ping rating scale and look forward to reading Volume 2.
!!!!!