Grid Computing for Developers: A Book Review

by Paul Barry

Title: Grid Computing for

Author: Vladimir Silva

Publisher: Charles River Media (Thomson
Delmar Learning)

ISBN: 1-58450-424-2

Price: US $49.95


Vladmimir Silva's Grid Computing for Developers
promises so much, offering to teach developers today's standards for
Grid computing, with the Globus Toolkit (GT) taking center stage. I
initially thought I was onto something very useful when I got the book
and was looking forward to learning more about grids. Soon, however, I
realized the book has a number of problems.

To start, the back cover refers to the book as "A Developer's Guide to
Grid Computing". It also states that the book is aimed at "beginner to
intermediate" level users, which it is not. After 50 pages of introductory
material, which sometimes is confusing and often is more marketing-speak
than technically useful information, a real shock came on page 55. That's
when the first source code is presented. The code is preceded by a short
disclaimer about how the code assumes the reader already is familiar with
parts of the GT, even though another 200 pages go by before the author
discusses those GT parts. Without that familiarity with GT, the code is
hard to follow. And, being that the book is aimed at developers, there
are pages upon pages of such code. Unfortunately, little accompanying
explanation is offered beyond brief overviews.

Beyond the problems with its content, the book has presentation problems
as well. It contains a number of glaringly obvious and embarrassing
spelling mistakes, the worst examples being "Baatch System" (page 119),
"Virtual zorganizations" (page 181), "Diffie and Helman" (page 193) and
"Web Ervices" (page 291). In addition, the included images mostly are
grainy bitmaps that look as though they were knocked together in a desktop
spreadsheet program. Screenshots are no better; they're grainy, poorly
labeled and sometimes incredibly small. For example, the screenshots
on page 260 are no bigger than a one-and-a-half inch square. And, the
less I say about the impossible to read let alone follow UML diagrams,
the better. Furthermore, the index is next to useless: try looking up
"OGSA-C", which appears on both the front and back covers but is not
listed in the index. Finally, being a computing text on a hot new topic, this
book is full of acronyms. But, no glossary and no list of acronyms are
provided. This book badly needs both.

In truth, I should have listened to the little warning that went off in
the back of my brain when I read on page 4 that the creation of
the ARPANET was one of the "milestones of the 1950s". Again, another
embarrassing error.

If you already are familiar with grids and the GT, there's probably
material in this book that is worth reading, although discussion of GT4
is thin on the ground. If you are new to grids, look elsewhere for a better

Paul Barry ( lectures at the Institute of
Technology, Carlow, in Ireland. Information on the courses he teaches,
in addition to the books and articles he has written, can be found on
his Web

Load Disqus comments