This book is vastly inferior to available alternatives. My primary complaint is that it appears written by an instructor familiar with various networking-related technology, but lacks a great deal of current real world design and implementation experience. My criticism exemplifies itself in what the book discusses, how concepts are explained, what is included, what is left out and the types of mistakes found throughout the book. While not the worst book for you, it is at best incomplete, while many times misleading, confusing and outdated at worst. Much of the material in the book is reasonably accurate, but there are a great many problems with this text that in total force me to advise students and learners of networking to avoid this title in lieu of other, far better alternatives. This review is based on the fourth edition published in 2010.
Forouzan's TCP/IP Protocol Suite is organized in a manner similar to many other general TCP/IP networking books. It begins with a history and background of the Internet and protocols in general. It then covers major protocols in the TCP/IP suite beginning from the network layer (IP, addressing, routing), through the transport layer (UDP, TCP, SCTP), the application layer (which is a mixture of DHCP, DNS, TELNET/SSH, FTP/TFTP, HTTP, email protocols, SNMP and multimedia), then back to the network layer with IPv6, and finally a few chapters on security, which are largely about cryptography services.
Perhaps like any text of this subject, the difficult part is deciding what to leave out and what to include. What is included here that would probably have been better left out or perhaps greatly revised include much of chapter 3, Underlying Technologies. This chapter is a summary of mostly layer 2 technologies including some IEEE LAN standards, Bluetooth and WAN technologies such as T1, frame relay and X.25. There is practically no attempt to relate any of this material to the TCP/IP protocol suite so it is largely superfluous and could have been left out entirely with little consequence. It might have been better written to demonstrate how these underlying technologies are used by IP, but no serious attempt is done so. In this chapter, chapter 7 (IPv4) and chapter 8 (ARP) ATM-related technologies are covered. Suffice to say, ATM is largely relegated to a niche technology at this point and at least in comparison to MPLS, which is more widespread, there is little reason to include so much about ATM if at all. The text also places far too much emphasis and includes too many examples on classful IP addressing, a long obsolete initial approach to IP addressing. In chapter 5 (IPv4 addresses), from page 121 to 135 classful addressing is discussed and in comparison the classless addressing scheme is covered only from page 135 to 146. Furthermore, about 20% of the end of chapter questions are about classful addressing. In chapter 12 (multicasting and multicast routing protocols), a sparse 3 short paragraphs are given to PIM-SM, by the far the dominant protocol used in the IP multicast-enabled Internet, while MOSPF and CBT, both obsolete protocols receive equal or more coverage. Chapter 25 includes a brief section on so-called Quality of Service ande short descriptions of network queue management techniques, but with no discussion on active queue management schemes such as RED and ECN. It further claims DiffServ is just a proposal by the IETF to replace the ToS byte, but has supplanted it for over a decade now.
What is left out of the text is more difficult to enumerate, but there is a clear sense that the author lacks a modern and practical knowledge of how real networks operate. The discussion of BGP is quite limited considering how critical it is to the routing subsystem of the Internet for instance. The security portions of the book focus largely on cryptography and are largely absent any discussion of DoS issues such as packet floods, TCP state exhaustion and amplification/reflection threats not to mention any real detailed discussion of firewalls and packet filtering mechanisms.
There are a number of very specific errors in the book. While many of them are of little consequence, they highlight at best a lackluster editing, review phase and carelessness of the author and at worst highlight the problems any serious student is going to run into when text and questions mislead. For instance, on page 266, question 16, the question implies that an ICMP message will be sent in response to a TCP segment delivered to a closed port, but no such message will be sent. TCP has a mechanism to respond to such misdirected segments, a TCP RST. The DNS discussion is outdated as the author seems to be completely unaware of EDNS0 and how widely implemented it is. One page 152, there are RFCs listed as having to do with IPv4 addressing, but some have nothing to do with addressing or were long ago made obsolete by newer RFCs. On page 131 example 5.14 claims 201.24.67.32 is a class B address when it is not and subsequently the network and mask given are incorrect. In question 10 on page 431, the decoded answer implies that a TFTP server will respond to a client on using a source port of 69, but in practice this doesn't happen as port 69 is only used for destination port in the initial request after which the server typically uses a ephemeral source port.
As a first course in TCP/IP networking, the approach the author takes to explain some topics are often not found in widespread use. For instance, the author covers a range of mathematical approaches including base 256 numbers and logarithms to explain IPv4 addressing. While mathematically there is nothing technically incorrect in doing so, some networking students without a strong mathematical background in my experience have found this author's treatment using this approach more confusing than helpful.
In my judgment the approximately $140 (US) cost for a new hard cover 4th edition of this book is vastly overpriced for the quality of what you get and especially so when compared to the alternatives. For a modern course in TCP/IP networking or as a reference to the TCP/IP protocol suite, other books available that do a comparable or better job may include Kurose and Ross' Computer Networking, Peterson and Davie's Computer Networks, or Comer's Internetworking with TCP/IP.
I had this book lying around with me for a very long time but never bothered reading it, because my graduate school textbook was something else. But when I eventually ended up reading it, i did regret how I missed out on this genius of a book, for so long. This book is way ahead of other books in its league. The material explained are illustrated very beautifully that there would be no place where you will find it difficult to follow the book. Fantastic book, i highly recommend it to anyone who want to understand TCP/IP.
If you are considering Comer Vol 1 or Richard stevens Vol 1 for TCP/IP , I would advise you give this book a serious thought too. Preferably goto a local bookstore to check them out and make a decision.
If you want to quickly learn/revise TCP/IP for an interview, this is a great book too because lots of times the self-explanatory illustrations saves you a lot of time from reading the material itself.
This may be the best textbook I have ever used. The text is concise and to the point with no extra "fluff". The illustrations are fantastic. The author has spent a tremendous amount of time on these... they take you directly to the meaning of the text, and give a strong visual and intuitive foundation to what has been written. Example problems with answers are numerous, and seem to be placed at exactly the points where you need to stop and work through concepts with pencil and paper. I really can't say enough about this text, I've read it cover to cover. It imparts an amazing amount of technical information without being dry... again, I think it's the illustrations that are so helpful, and it's nice that the author's style is direct and not wordy. I don't think you can buy a better TCP/IP book, this one's worth twice the price.
I had to use this book for a graduate course. It has been very suitable for this purpose. The level of details is good, but not to the point of a protocol standard. If you need to master the concepts and do not want to refer to very superficial practical references or detailed standards this book is right in the middle. Lecturers will find it very appropriate and students will easly grasp the concepts. One can still use some parts of the book as a reference. If the required information is not there some of the references to RFCs and standards will guide the reader to more information elsewhere.
TCP/IP is the language of the Internet and this book is its grammer. Although I think this book was intended primarily to be a college textbook, I'm not in school. I just was assigned a project to do some software requiring a direct interface to TCP/IP. I found this book to be exactly what I needed.
The book uses a very visual approach that allows you to quickly find the subject areas you need. Then the key points are put into boxes, kind of like someone has gone through with a hi-lighter to emphasize the critical points. These enabled me to skip rapidly through the points I knew about to get to the areas where I needed detail.
This book further is one of the best written and more clearly describes this rather technical subject in clear English words that even a simple programmer can understand.
This is the third edition, it covers some of the newer protocols like SCTP. Exceeding well done book.
Product Details :
Hardcover: 928 pages
Publisher: McGraw-Hill Science/Engineering/Math; 4 edition (March 25, 2009)
Language: English
ISBN-10: 0073376043
ISBN-13: 978-0073376042
Product Dimensions: 7.7 x 1.7 x 9.3 inches
More Details about TCP/IP Protocol Suite Mcgraw-Hill Forouzan Networking 4th edition
or
Download TCP/IP Protocol Suite Mcgraw-Hill Forouzan Networking 4th edition PDF Ebook
No comments:
Post a Comment