If you are reasonably familiar with the rudiments of computer networking and are unable to cope with the complexity and diversity of networking products inundating the market, if you feel that the time has come to bite the bullet and really understand at least one networking product in all its grandiose glory and diminutive detail, and if you do not mind if that product is from IBM, then this book is for you. Certainly, this is not a textbook suitable for classroom use: there are no examples, exercises, or references. Nevertheless, this is definitely a useful reference book or companion to another textbook.
As the title succinctly points out, the book deals with IBM’s solution to a networking problem. The reader is expected to know what this problem is; only then does the utility of the book become apparent. The authors do allude to the problem in their preface: “Networks are being built at a rapidly increasing rate, and these networks are of increasing complexity and diversity . . . [this] proliferation of devices, however, can cause compatibility problems. . . . In order to facilitate compatibility, hardware manufacturers developed network architectures. . . .” DEC’s DECnet, Xerox’s XNA, and IBM’s SNA are examples of such architectures. Academics and textbooks tend to focus their attention on either a theoretical analysis of a concept or a comparative study of the general features of competitive designs. Martin and Chapman skip these formalities and plunge right into the details of the design and implementation of a specific system.
The book’s 380 pages are organized into 28 chapters. An essentially top-down approach is taken in presenting the material. The first seven chapters deal with the basic concepts associated with SNA architecture, including one chapter on how SNA stacks up against the OSI model. The next five chapters deal with advanced facilities such as program-to-program communication, logical unit design, and distribution services. The following six chapters deal with path control; three of these chapters contain a very nice description of SDLC. The remainder of the book goes into a detailed description of transmission and data flow control and function management of the network addressable units (NAUs). The presentation throughout is concise and clear. Not a single equation can be found in the entire book. Like many other James Martin books, this one is profusely illustrated, often in color. As a person who has spent equal time on both sides of the industry/academia fence, I feel that this book contains valuable practical information that academics often tend to ignore in their pristine pursuit of equations. The book would have been better, though, if a chapter had been devoted to a comparison of SNA with another network architecture.
SNA has gone through several evolutionary stages since its inception in the early seventies. Without a doubt, it will continue to evolve. The reader should be aware of these developments to keep really abreast of SNA.