-- Leo's gemini proxy

-- Connecting to dfdn.info:1965...

-- Connected

-- Sending request

-- Meta line: 20 text/gemini;lang=en-US

Networking Books


I have purchased dozens of networking books over the years and have noticed that some books are an absolute waste of money. These books are what I call "exploitation books". They take advantage of the explosion in the number of people who need to learn networking. These books tend to be sloppily written, and are sometimes no better than the RFCs in terms of readability.


I prefer books that take the time to explain concepts in easy to understand language and offer examples. Books that claim to be for beginners but which assume you are already an expert are a waste of money. Also in the books-to-avoid category are books that do not deliver on the claims made on the back cover. The majority of these were purchased through book clubs sight-unseen.


Books To Avoid: These don't deliver what the advertising claims


Must Have Books: Essential for any networking professional


Books Worth Purchasing: Not essential but nice to know information


Books To Avoid:


The Taylor Networking Series

The Taylor Networking Series, written by D. Edgar (aka Ed) Taylor, and published by McGraw-Hill is made up of five separate books. They include the following:

Multiplatform Network Management, copyright 1997

297 pages, $55.00


TCP/IP Complete, copyright 1995

300 pages, $60.00


The McGraw-Hill Internetworking Handbook, copyright 1995

773 pages, $60.00


The Network Architecture Design Handbook, copyright 1998

764 pages, $60.00


Encyclopedia of Network Blueprints, copyright 1998

1,122 pages, $60.00


Avoid this entire series of books unless you work in an IBM shop or you have money to burn. In general most McGraw-Hill books are barely acceptable in terms of readability and usefulness, but these particular books are just plain useless. Part of the reason is that the last three books listed above are essentially the same book with a different title. Taylor apparently thought that no one would buy all three of these books and would therefore not notice that he reuses the same material in each book, right down to the figures and tables.


The Encyclopedia of Network Blueprints claims on the back cover to be a book "Covering all the bases in networking from layout to tweaking...". That is simply not the case. The "blueprints" are actually childish diagrams that are not much better than the network designs you scrawl on napkins at lunch. The eight TCP/IP "blueprints" are essentially identical and ambiguous. They are not actually "blueprints", but rather some incredibly sloppy drawings of networks. Incredibly, Taylor does not mention subnetting in any of these TCP/IP "blueprints". This book was sold using some incredibly false advertising since it does not even approach the level of information the advertising claims it contains.


The cover specifically indicates the book contains "blueprints" for ATM, Gigabit Ethernet, and Frame Relay. The only mention of Gigabit Ethernet is found on page 39 and is limited to a 35 word paragraph. The section on ATM is limited to 28 pages. The section on Frame Relay is 14 pages in length. Keep in mind that the book contains 1,123 pages.


This book includes all of the protocol specifications found (verbatim) in Taylor's other books. No effort is made to explain any of it. It is just a repeat (and probably a cut-and-paste) of the RFCs. This technique is what makes these three books appear so big.


He does not cover switching at all, even in the "section" (the 35 word paragraph) on Gigabit Ethernet. He apparently has never worked with Layer 2 switches which explains why a book that purports to be everything you need to construct any type of network does not mention Ethernet switches.


In an effort to make the book bigger, Taylor includes over 100 pages of useless system information from his network. These are simply print-outs of diagnostic programs found on Windows NT 4.0. There is absolutely nothing useful to anyone to be found in these listings. They simply add to the number of pages in the book, possibly in an attempt to make the book longer as a way to justify its price.


The most thorough sections of the book, and even these are inadequate, cover the designs which you would find in an IBM shop, including APPN, SNA, and believe it or not, VTAM (that is an oldie). This may be because Taylor worked for IBM. Taylor actually wastes space with detailed descriptions of IBM accounting machines introduced in 1952 such as the IBM 701, and systems from the 1960s such as the System/360. Again, this technique seems to be a deliberate attempt to increase the size of the book.


Tables in the book are misnamed and filled with useless, incomplete information. An example is a table which purports to list "Routers, Hubs, and Switches", but which actually lists IBM monitors and keyboards. Another is a table which has the heading "Storage Devices" but which contains part numbers for IBM rack enclosures. Perhaps as a way to aid his former employer, IBM, Taylor includes mountains of IBM product information that is completely useless, but which does add to the page count.


The section on ATM is basically a cut-and-paste of the relevent RFCs. The ATM "blueprints" are not something from which you could actually build a network. All the work you would assume the author did in preparing the "blueprints"- like equipment selection- is something you will have to do yourself. Because the book does not include any mechanism to get the "blueprint" files (e.g., a web site or accompanying CD) you would need to create these manually. And the diagrams are so vague (e.g., "ATM Switch", "ATM Concentrator") that you will have to start from scratch anyway. In the sections on IBM networking Taylor lists equipment by model number, but in all other sections he does not mention any specific manufacturer or model number for routers and hubs.


As you proceed from front cover to back cover in "Encyclopedia" you notice that the level of detail gets lower and lower. This may be because Taylor was under a deadline to get the book to McGraw-Hill. Since he apparently spent a lot of time on the technologies he knows about (all IBM related), there was no time left to learn and document the technologies he obviously does not know (ATM, Frame Relay, Ethernet, subnetting, Layer-2 Switching, and Layer-3 switching).


Avoid all of these books. The large investment can be better spent almost anywhere else.


LAN Blueprints

Gerald T. Charles, Jr.

McGraw-Hill

Copyright 1997

208 pages

Two diskettes

$44.95

Talk about useless books. This one appears to be a collection of files that the author used at some point in the design of a network. It is not clear what kind of network it was, because he doesn't spend any time with explanations. Instead the book is made up of the various documents a consultant provides to a customer. The diskettes contain Excel spreadsheets and bizarre, completely nonsensical Microsoft Project Files. Alas, the network diagrams were created by some obscure software called "Canvas" which produces images files with an extension of .CVS. I have not found any software that will convert (or even recognize) this software. Apparently the Canvas software has the option of saving files in .PCX format, but the author choose not to save the files in that format, probably thinking that purchasers of his book would run out and buy Canvas for the sole purpose of viewing his network diagrams. This is a pretty good indicator of the lack of planning that produced this book. The author's use of this software is also a good indicator of just how out of date the material in the book is. Although the copyright is 1997, the material is probably from around 1994. That explains why the emphasis is on LANs, with only a pasted-in mention of WANs. The cover of the book says "50 Blueprints" but I could find only one blueprint (if you can call it that, I don't). Perhaps the cover is referring to the 50 files on the accompanying diskettes. Avoid this wildly overpriced book at all costs. This book does not even come close to fulfilling the advertising on the back cover. It is another excellent example of the exploitive nature that surrounds the networking book business these days. It is no coincidence that this is another McGraw-Hill book.


Essential Windows NT System Administration

Aeleen Frisch

O'Reilly and Associates

Copyright 1998

208 pages

$34.95

The author is a UNIX administrator who has written other books for O'Reilly, including Essential System Administration, a book for UNIX administrators. The Windows NT version is a flat-out waste of money. This is an example of an author trading on previous works to sell more books. This is one of the few O'Reilly books that is not worth the paper it is printed on.


Must Have Books:


TCP/IP Illustrated Volume 1, W. Richard Stevens

Copyright 1994, 576 pages, $45.00

Simply the best book on TCP/IP. Written in a clear, easy to understand way, this book guides you through the complexities of the two protocols.


Getting Connected To The Internet at 56kb and Up

Kevin Dowd

Copyright 1996

O'Reilly and Associates

410 pages

$29.95

Explains, in easy to understand language, practically everything you need to know to get a network connected to the Internet. Excellent sections on Frame Relay and router configuration. A must-have if you are new to getting connected.


Voice & Data Communications Handbook Bud Bates, Donald Gregory

Copyright 1996

McGraw-Hill

699 pages

$65.00

Excellent explanations of T1 signaling and framing, in-band, in-channel, out-of-band, and out-of-channel (common channel) signaling.


Newton's Telecom Dictionary 11th Edition

Harry Newton

Copyright 1996

Flatiron Publishing

685 pages

$34.00

Thorough and often humorous explanations for the thousands of buzz-words in the telecom industry.


ISDN and Broadband ISDN with Frame Relay and ATM 3rd Edition,

William Stallings

Copyright 1998

Prentice Hall

581 pages

$64.00

Includes an excellent chapter on Signaling System 7 (SS7) which is important if you are working with voice over IP (VOIP). Also includes excellent explanations for Common Channel Signaling. The chapter on digital transmissions is very well written and provides a wealth of information on the local loop.


DNS and BIND 2nd Edition

Paul Albitz and Cricket Liu

Copyright 1997

O'Reilly and Associates

418 pages

$32.95

The definitive work on DNS. Easy to read with plenty of examples, this book walks you through actually setting up a DNS server. Keep this one close at hand if you will be setting up and maintaining DNS servers. Well worth the money.


Managing IP Networks with Cisco Routers

Scott M. Ballew

Copyright 1997

O'Reilly and Associates

334 pages

$29.95

Includes specifc examples for configuring interfaces, Frame Relay, and routing protocols. Helps get you through the Cisco documentation.


Books Worth Purchasing:


SNMP, SNMPv2, and RMON Practical Network Management 2nd Edition,

William Stallings

Copyright 1998

Addison-Wesley

478 pages

$65.00

This book is not essential, even if you are using SNMP and RMON, unless you are writing software that uses either of these management protocols. It is probably overkill for network managers, but does offer the most complete information available.

-- Response ended

-- Page fetched on Mon May 13 16:51:43 2024