Enterprise Integration : The Essential Guide to Integration Solutions (Addison-Wesley Information Technology Series)

Rating:
Summary: Elegant, Easy to Read, Good Primer for Managers
Review: If you are a manager to whom information technologists report, or a manager that employs technical advisors who in turn help oversee varied IT procurements and implementations, then this book is an ideal primer. It can also be scary, because I will wager than in 7 out of 10 cases, the technical experts are not pursuing the enterprise integration fundamentals that this book outlines.

Both authors are strong in their own right. The book bring together Bill Ruh, former MITRE, MITRETEK, and Concepts 5 guru, today the global manager for CISCO AONS, who is updating his 2000 book on the topic, with Beth Gold-Bernstein, who has consulted, lectured, and written on this topic, and has her own book titled "Enterprise Integration: A Practical Approach."

I regard the book, and the topic, as a watershed between the old days of configuration management and a focus on data that was largely within internal custody, and today, when real-time data integration and exploitation is required across both all internal points (i.e. including the 85% that is in emails and hard drives) and external points--not just the web, but supplier, buyer, regulatory, and other databases.

I recommend this book for managers in part because the book itself is quite clear on the fact that information technology by itself, no matter how much money is thrown at it, will not achieve enterprise information integration. Management mind-sets, management metrics, management enforcement of standards and compliance with the strategic direction implied by enterprise integration, are all required.

Early in the book there are important references to both scale and speed, with the key difference between the 1990's and today being that instead of humans accessing the data, there now much more machine to machine communication and sharing, and this requires hyper-speed. There is also much more focus on event-driven information actions, with Delta Airlines being cited as a very good case study--the system must be able to take many autonomous actions triggered by an event (e.g. an airplane more than 15 minutes late, with repercussions across gate management, luggage management, connections management, catering management, etc.). Zero latency, real-time enterprise, and event-driven information transactions are among the buzz words.

The case study of CISCO on page 6 grabbed me early on--my primary focus is on the Global War on Terror (GWOT), and reading about CISCO's move to real-time metrics (this book is *very* strong on metrics, which I take to be a very good thing) and real-time decision making and course corrections, I was thinking to myself that CISCO is to information as special operations are to terror. So when CISCO doubled productivity, cut costs by 30%, and made daily reporting the norm, I say to myself: okay, now let's see that in GWOT....this book is Ref A in answering that challenge. Another case study, on FedEx using hand-held devices as both points of data entry in the field, and end points for data value to the field, also struck me as relevant to GWOT.

Throughout the book, one of its own phrases: "people are the most expensive part of any system," keeps resonating, because everything in here is about either increasing productivity or reducing the time-cost of information transactions. This book also has a very healthy focus on information sharing across all boundaries, with appropriate security, privacy, and legal attributes for each transaction.

Standards receive heavy emphasis throughout.

The book is slightly dated on the topic of automated metastandards and semantic data definitions, but I know the authors to be personally very engaged in the very latest developments surrounding semantic web and synthetic information architectures and other related automated assignments of meaning, so I take this to be primarily an issue of timing--the book had to be put to bed.

The chapters on Information Integration Architecture and on Information Integration, the ones I was most looking forward to reading, strike me as the least developed among the many excellent parts of this book. In part this is because Enterprise Content Management (ECM) is just coming of age, and truly scalable solutions to the challenge of managing global multi-media multi-lingual unstructured information data (Cf. InfoSphere AB in Sweden) are just now coming into being. This chapter does provide an important itemization of key organizations responsible for metadata standards, and lays out a framework for establishing "who needs to know what when" as part of the manager's contribution to the over-all enterprise integration planning process. These two chapters excel in pointing out that information management is about ensuring long-term data value, allowing for reachback over time and space.

In its conclusion the book makes reference to turf wars, training, reducing redundancy, reducing reliance on proprietary technologies with lock-in costs, finding a return on assets, and creating a culture of reuse. The last hundred pages of the book, and the CD-ROM, provide templates that any manager could reasonably demand of their technical advisors. I opened these up and found them very useful, to the point of being worth at least a week if not more of man-time, and hence easily repaying the price of the book many times over.

The bibliography is good and the index has been thoughtfully developed. I recommend this book to anyone who deals with global information in any form, but especially to managers who might be wondering if their IT people have any clue as to where they are taking the enterprise and its information. This book also strikes me a superb textbook, both for undergraduates as a primer, and for graduates as a foundation for a more nuanced discussion. For myself, it was "just enough, just in time" information, exactly what I wanted and needed in my specific context.

Rating:
Summary: Great book on a timely topic!
Review: Nothing is hotter in IT right now than integration. The authors have done a great job in not only explaining the term "enterprise integration," but arguing that it's as much a business concept as it is a technological one. The short case-study examples propel the text forward and give it a real-world credibility. The authors have obviously walked the integration walk, and their book is an effective testimonial to their experience.

Jill Dyche
Author of The CRM Handbook and eData

Rating:
Summary: Essential templates to integration
Review: The statement on the last page, "This book is not intended to be a full blown integration methodology", holds true. This book is essentially a walkthrough of the provided document templates (see Appendix and the included CD) and provides only limited information about how to use or implement an integration methodology. In fact, the templates are the main reason to buy the book, but don't expect to read about a tried and tested integration methodology or (unfortunately) how to develop an enterprise integration architecture.

The authors do include a fair number of 'lessons learned', but they are scattered across a very repetitive book, so don't assume you can easily find them again. I found the authors' choice of 'important lessons' odd at times. For example, section 8.3 makes some key points about metadata for integration architecture, but the points only feature in the standard text whereas a description about XML is highlighted in a framed box. That really seemed like a waste of two pages.

The second key part of the book is the "Integration Road Map" first introduced on page 11. The road map (not to be confused with a methodology!) is meant as a "step-by-step" guide to implement a reference framework based on the provided templates. Incidentally, it also serves as a reading guide to the book.

The problem with the road map is that it fails to explain how the individual activities (read: templates) hands together, i.e., they lack an obvious way to link the templates together into a coherent architecture description. This is why the book falls somewhat short of the stated goal of demonstrating how to document a reference integration architecture. 'Disjointed' is the word I was looking for.

The authors' have focused on what the templates should describe rather than how the templates describe a certain view or aspect of the enterprise integration to ensure a consistent end-to-end architecture. "Documenting Software Architectures: Views and Beyond" is a good reference on how to do this.

Oh, and before I finish. The book's constant reference to importance of building "re-useable" components is not a bad thing, but just remember that good re-useable software components only come from the knowledge of what and how people will re-use them - guessing will almost certainly only lead to wasted development effort and undermine the business' confidence in the IT department or vendor.