On Demand Computing: Technologies and Strategies

Being that I specialize in IBM software as part of my IT job, I figured that reviewing On Demand Computing - Technologies and Strategies by Craig Fellenstein (IBM Press) was a good choice. After finishing my read, I'm somewhat torn on how to rate it... Chapter List: Part 1 - On Demand Business: Introduction to IBM On Demand Business; The On Demand Operating Environment Part 2 - Autonomic and Grid Computing: Autonomic Computing Strategy Perspectives; Grid Computing; The Future of Grid Computing; Grid Computing Strategy Perspectives Part 3 - Service Providers and Customer Profiles: The On Demand Business Service Provider Ecosystem; Industry Matters and Customer Profiles; Conclusions Appendix A - IBM On Demand Developers Conference; Glossary; Reference Materials; Acknowledgments; Index First, the positives... IBM has committed to building autonomic systems; systems that monitor, adapt, and heal themselves. Outside of this book, I don't think there are any other sources that cover the subject as much as this one does. The material on grid computing is very well done. Not only is there the technical foundation for how it works, but there's a fair amount of detail as to how it is implemented. People charged with investigating and building these types of systems would be able to gain a great deal of understanding. The material on case studies is rather interesting too. There are a number of short pieces on companies who have changed their business processes by building/buying IBM systems that are classified as "on demand". Now the flip side... I had a real tough time wading through parts of this book. It could be that I tend to be more practical and hands-on. The material on autonomic computing seemed to be a lot of very high-level concepts repeated over and over. I kept thinking "didn't I just read that a couple of pages ago?" Either the details weren't there, or I was glazed over and missed them. I felt as if I was hearing concepts with no solid factual examples of how "on demand" is implemented. Even the case studies at the end could, in some cases, be seen as business process reengineering and not necessarily anything radically different than other businesses that have undergone major IT initiatives. It just so happens that they used IBM systems, hardware, and resources, and I felt as if some of the cases were labelled "on demand" on that basis alone. I'll give this the benefit of the doubt and assume that some of the problems might have been me. If you want to know more about On Demand or grid computing, this is definitely the book to read. I'll just warn you that some parts seem to be more idea-based than reality-based...

This book was a very easy read of a very complex topic. Well structured for anyone to understand. The author effectively describes the basic characteristics of an onDemand system and why we need it. This book changes onDemand from marketing hype and enables the reader to understand why this is a fundamental model of computing necessary for the future.

IBM continues to build out its vision of the future direction of computing, as exemplified in this book by Fellenstein. It builds upon earlier discussions on grid and autonomic computing. Fellenstein shows how those ideas can be extended into a more complex realm, that IBM calls On Demand Computing. While the term is simple, he goes to some lengths to show that this is not something as mundane as the common idea of utility services. Unfortunately, that may be the typical reaction to the term, so he has to do this. If you go thru this book and see the meat of the discussion, you'll realise that it includes complicated technologies that are barely built. Like the autonomic approach, which borrows metaphors from biological systems, like monitoring and prevention. But even beyond these are what the author calls ecosystem dynamics. A system of disparate hardware and software interacting across a network, wired or wireless or, probably most commonly, using both modes. Somehow, this system has intelligence at its nodes, to ask for or supply computing resources. Even IBM can't do it all. This approach also has the merit to IBM of shifting it from supplying hardware or operating systems, both of which may be moving to commodity status. Having said this, the book could have well been trimmed in its earlier chapters. A fair amount of repetitive material.

In the style of his previous books, this new book by Mr. Fellenstein, describes "On Demand Computing Technologies and Strategies" in clear and easy to understand terms. This unique book discusses many of the underpinnings of the respective technologies, as well as, provides a full treatment of the strategies-including many case studies. Mr. Fellenstein describes in a very eloquent way, many of the ambiguous terms in the industry related to the topics surrounding "on demand." He does this in a global sense, while including perspectives from different geographies in the world. This book is well organized, easy to follow and understand-a great blueprint for how global technologies, strategies, and standards have evolved-and will continue to evolve. I highly recommend this book, and congratulate Mr. Fellenstein for a job well done on bringing to the global markets-this unique and very informative book.