This book has been reviewed many many times before, and because it is such a brilliant masterpiece, it thought it was worth while mentioning this again.
I simply don't have words for this book. Everything is so well explained, well documented, well written... it simply is a must read for anybody who wants to do anything with BGP.
Personally, I have found it valuable to first learn some basics about BGP before reading this book. But even if you are a BGP beginner, you will like this book.
High five, well deserved !
This is a great book to get the fundamentals of BGP and many advanced concepts related to BGP on the internet and in your enterprise. Even though Halabi wrote the latest edition almost 10 years ago the information is very relative to today. It is of course very Cisco specific but gives you a great understanding of how the protocol works and will help any engineer who deals with BGP. A must read.
I have always felt that I am part of the small minority that just does not see the beauty of this book if studying for the CCIE. I have, for years, read other's comments on the how Halabi's "Internet Routing Arch" is a `must have', yet I still have not found a similar reason why. While I do feel that the book does (probably) the best job at explaining BGP and how to implement BGP in a Cisco-centric world, I have not discovered the value for the CCIE. I would much rather spend my time with CiscoPress's "Routing TCP/IP" Volume 2, by Jeff Doyle and Jennifer DeHaven Carroll than this book.
One item of concern, there is a noticable amount of trivial errors in this book that can lead to disaster if implemented incorrectly. For example:
Page 106 - "...any traffic that has an origin OR destination that does not belong to the local AS." This clearly should be an AND, not OR.
Page 315 - "You can also specify a PREFIX list..." - I believe this should be a filter list, not a prefix list.
And then there are places where the book is too opinionated - for example - page 206 "Many operators choose to filter dynamically learned defaults to avoid situations in which traffic ends up where it is not supposed to be." I do not feel this is true, and can think of multiple times when not filtering the advertised default route may just end up being the worst option. In fact, I do not feel either configuration is the right configuration to admit or condone, as the wrong configuration can have disastrous effects.
I still have to give this book 4 stars - simply because of it's utilitarian value and overwhelming sense of loyalty given to this book by others. But I don't reach for this book too often.
Internet Routing Architectures is considered the BGP4 handbook and rightfully so. I found the overall composition easy to read. Even though the book is aimed at varying levels of expertise, it does not assume any level of knowledge in TCP/IP or routing. It explains all the concepts, from the simplest one right up to the most demanding, in a fluent way. The book works hard not to withhold protocol details and design-oriented information, while at the same time realizing that building general understanding comes first. An in depth look at BGP is supplied here from the theory aspect, if you need actual configurations and commands I suggest also:
Cisco BGP-4 Command and Configuration Handbook : CCIE Professional Development
I've read both editions of Halabi's book cover-to-cover and I was stunned to discover just how widely they expanded the excellent original material. Though the configuration examples are IOS-centric, you will have absolute mastery of BGP routing and its various topologies by the time you finish this book. You'll learn the particulars of redundancy and load balancing, with plenty of concrete examples that you can apply directly in your own networks (I did), how to configure Multihop and when you may need to do so, how to set up multihoming in all its permutations, filtering and route maps, and much much more, in just about every possible deployment. Among the advanced topics, you'll understand how confederations and route reflection work, and how confederations help consolidate large BGP topologies into more manageable units - and also their limitations. It's basically impossible for me to lay out every topic that is effectively discussed in this book.
If I had to choose one book for core networking topics, this would probably be the one.
mr halabi knows his material and has an easy-to-follow writing
style. however, this book does an inadequate job of providing
complete examples (especially community usage), which when
learning something as complex as bgp, are essential.
overall, it is a great book and i recommend it to experienced as
well as novice engineers, if for no other reason than the
discussion of some of the politics and history of internetwork
engineering (as well as some of the organizations/groups which
exist)
this book will teach you cisco's version of bgp - but when you
have cisco's marketshare, why not... you should alse read the
rfcs for bgp. before you think about trying to set-up bgp read
this, but also do some searching on the web for other resources
(the multihoming faq is very useful)
After a brief review of the history and evolution of the Internet, the author motivates the rest of the book by giving a set of questions to be asked by an organization who intends to connect to the Internet. Since at this time all businesses it seems want to do this, the answering of these questions will give them more helpful information on just what must be considered when choosing an ISP. The details of the routing architecture are of course the main emphasis in the book.
Some of the more important topics addressed in the book are: 1. IP subnetting and variable-length subnet masks and why they are useful. 2. The different strategies used to handle IP address space depletion, such as creative IP address allocation, classless interdomain routing (CIDR), private IP addressing, and the new IP version 6. In the discussion on CIDR, the author asserts the advantages of using aggregation, in that an Internet Service Provider can advertise one IP network rather than several individual advertisements. This, he says, results in more efficient routing strategies and propgation along with making the route advertisements more stable. The degree of the resulting efficiency is not really quantified by the author however. It would have been interesting to have real-life examples of the resulting gains, or examples taken from simulation modeling. Although such data might seem unnecessary now, since CIDR was proposed as a fix to the depletion problem, it would still be interesting to be able to understand in more detail the advantages of employing CIDR, and with comparing it the planned deployment of IPv6. 3. The discussion of distance vector routing and link-state routing and the advantages and disadvantages between the two. The reader interested in a more rigorous and quantitative comparision between the two routing protocols will not find it here, but such a comparison can be done via simulation modeling. 4. The representation of the BGP neighbor negotiation via a finite state machine. 5. The discussion of the TCP MD5 Signature Option, and its role in protecting BGP from spoofed TCP segments and TCP resets. 6. The building of peer sessions using BGP and how to implement it "internally" in an 'autonomous system.' Peer connections between routers in different autonomous systems are then "external" implementations of BGP. The autonomous systems as explained by the author can be used for example by two users who desire to have a link between them in case of a failure of their ISP. 7. The discussion on route instability and how to control it using aggregation, route dampening, and static route injection. The author spends an entire chapter in fact on the design of stable internets, although the discussion is brief and purely descriptive. The route dampening mechanism is discussed as a tool for controlling route instability. This involves a strategy for penalizing unstable routes and is implemented on CISCO routers (the author gives the commands for doing so explicitly). Although the author does not discuss any, modeling and empirical studies have indicated that a cavalier use of route dampening can be deleterious to a network. For example, it was shown early on in the use of BGP that a single route withdrawal may cause other routers to explore a sequence of alternative paths before deciding that the destination is unreachable. Some researchers have shown that this in turn causes "secondary flaps" which can suppress the threshold of the route flap damping algorithm, and resulting in delayed convergence of the route.
I am the architect of a leading MPLS VPN service which uses BGPv4 with MPLS extensions. I constantly refer back to this book for standard BGP scenarios (it has no MPLS VPN info). I have recommended this book to all our techincal support people and our customers for designing BGP peering networks.
The book is well written, well organized and easy to follow. It has great breath of BGP applications. It is a tremendous help for those designing AS peering networks.
It gives a brief overview of internet routing then dives into BGP. After the BGP protocol description, the next part of the book has applications geared around scenarios/case studies. The last part is specific Cisco IOS configurations for the scenarios in the previous part. The scenarios cover, to list a few: load balancing, preference routing (primary/backup), route redistribution, default routes, route summarization, route reflectors and confederations (scaling), damping (stability), policy control (filtering/manipulating routes, attributes and community values).
Note that it is not a BGP specification nor a Cisco IOS reference. A better title would be "BGP Applications".
Product Details :
Hardcover: 528 pages
Publisher: Cisco Press; 2 edition (September 2, 2000)
Language: English
ISBN-10: 157870233X
ISBN-13: 978-1578702336
Product Dimensions: 7.6 x 1.5 x 9.4 inches
More Details about Internet Routing Architectures 2nd edition
or
Download Internet Routing Architectures 2nd edition PDF Ebook
No comments:
Post a Comment