Title "Network Performance Baselining" Author Daniel J. Nassar Publisher Macmillan Technical Publishing Publication Date May 2000 ISBN 1-57870-240-2 Price $50 (U.S.) Pros A well-balanced overview on the topic, focused on the expert public, included case studies, and was vendor-neutral. Cons Technical errors prevalent, new developments not included, failed to follow through on in-depth issues. Verdict If you have an urgent need to start network performance baselining tomorrow, get this book. Otherwise, it will probably pay to wait for the second, updated edition.

The book gives a good overview of some of the critical issues in measuring and documenting network performance, and presents this information in a few brief, informative chapters. It also provides an overview of baselining tools, such as protocol analyzers and specific monitoring tools. The book also provides a vendor-neutral, readable synopsis of some networking history, focusing on troubleshooting performance issues. There are chapters on LAN and WAN protocols, Ethernet, Token Ring, Fiber Data Distributed Interface, ATM and WAN. The general chapters about measuring and documenting network performance issues are the most valuable contribution this book will make to your library.

A less attractive feature of the book is the large number of technical errors in the specific technology chapters, which could easily baffle non-experts in the field. It's for this reason that I would not recommend the book as a resource for beginners. For instance, in its overview of the TCP/IP stack, the book places the Address Resolution Protocol (ARP) on the logical link layer (fig 7.14) and 20 pages later (Fig 7.32) the ARP appears in the network layer where it belongs. Similarly, on page 306 the book states: "When examining UDP, keep in mind that the UDP provides only the function of assigning and allowing access to a port for TCP-based communications." While I suspect that the author means IP-based communications, it's statements like these that make this book unsuitable for all but seasoned experts.

I then wonder how much these seasoned experts will get out of this book. The tools and methods that are discussed all seem a bit dated, and many important recent developments are either not discussed at all (multi-layer switching, for instance) or only barely touched. The author mentioned RMON as a management tool for switches, but only mentions that there is RMON 1 and RMON 2. He doesn't mention that RMON is specified in RFC 1757 and allows for the collection of statistics, history, alarms and events. In a book dedicated to performance baselining, this should have been discussed extensively.

Finally, a book with the coverage of this one should have an extensive list of further resources for technical details and documentation. This book is seriously deficient in this area. It assumes that you know where to find these resources yourself. A reference to "the Microsoft Web site" for instance, is only minimally helpful. I would also expect that Cisco's site (www.cisco.com) and the IETF (www.ietf.org) as sources for specific documentation would have been worth a mention.

I really hope Daniel Nassar will consider expanding on this work in a second edition with more current coverage, less technical errors and a better resource list. Such a book would be well worth having. In the meantime, if you have an urgent need to start network performance baselining tomorrow, by all means go ahead and get this book. If your need is less urgent, it will probably pay to wait for the second, updated edition.

Have you read this book? Let us know what you think! Rate it below or enter our Forums.

More articles by Hinne Hettema:

• It's All in There
  
• Clustering 101
  
• My Top 10 Study Tips for Passing the Designing Security Exam (70-220)
  
• ADSI Scripting for Pros