This issue of CBR nicely follows our September issue on best practices in enterprise systems. Both the ES and SOA trends have as one of their principal drivers the need for integration. They approach it very differently, of course, but they both cut across the organization and therefore engender similar dynamics and problems - both technical and organizational - that must be proactively managed. (Hey, here is an idea! Maybe we should begin to create themed CBR collections by bringing together and reissuing installments with similar or related focus. OK, I'll pass this idea on to Cutter headquarters ... ) In the meantime, I hope that you will find the frameworks, ideas, data, and guidelines in this issue of CBR to be helpful in your quest to evaluate and possibly implement a service-oriented architecture in your organization.
What is it about humans? It seems that we are constantly striving to do more, get more, invent a better mouse trap. If you think about it, we may very well be the only animal in the kingdom that has this peculiarity. Add to this the ability to communicate and record our progress, and you have a recipe for continuous innovation and advancement. This hyperactive learning and innovative trait is apparent in few other endeavors as much as it is in technology - and information technology is the ADHD child in the technology family!
Of course, the more cynical amongst us can say that it is just vendors and consulting firms harping "new" stuff (typically, old wine in new bottles) to make a buck. As always, the truth is probably in between: new technologies hyped as breakthrough may be just partly new, but they are novel nonetheless and they do present some new challenges as well as some old risks.
It is for this reason that this issue of Cutter Benchmark Review focuses on one of the hot issues in information technology today: service-oriented architecture (SOA). We do so with our tried-and-true CBR approach: no rhetoric, no hype, and a solid understanding of "today's new thing," based on the decades of experience of our contributors in the fast-moving IT world.
SOA, of course, is new in part, but it is built upon a tradition of technology and a progression of business needs. On the technology front, SOA is the latest in a series of technologies focusing on reusable code and modular designs: objects, components, enterprise application integration. On the business side, the need for SOA has been building with the increasingly pressing data and application integration imperative and the more recent focus on strategic agility and flexibility.
With a topic like SOA, where the hype potential (i.e., hype danger) is significant, CBR's approach is particularly valuable. Our faithful readers know what we do: We invite two contributions from individuals with significant experience creating, researching, and/or managing IT. One of our contributions comes from the academic world, while the other comes from the trenches of day-to-day IT consulting and management. Our contributors frame the topic, tell us what we know and don't know about it, explore how it relates to previous trends, detail the pressures and interests surrounding it, and typically offer a framework for us to use in making sense of the topic. They also analyze the results of a Cutter survey on the issue at hand and draw actionable guidelines for our readers. Their considerable experience, rigor, and "having been around the block a few times" come in very handy when we tackle a complex, multifaceted trend like SOA.
Let's do it then!
Our academic contribution in this installment of CBR is provided by Andrew Schwartz, Assistant Professor of Information Systems in the EJ Ourso College of Business at Louisiana State University (LSU), and Rudy Hirschheim, Ourso Family Distinguished Professor of Information Systems in the EJ Ourso College of Business at LSU. Both Andy and Rudy focus on the managerial and organizational aspects of new information technology and are becoming increasingly interested in SOA. Providing our view from the field is Michael Rosen, Director of Cutter Consortium's Enterprise Architecture practice and a Senior Consultant with its Business-IT Strategies practice. Mike was the Chief Enterprise Architect at IONA Technologies, PLC, and has decades of experience as an application architect, designer, and developer.
Andy and Rudy first define SOA and then, drawing on their considerable experience researching enterprise systems (ES) and the challenges they engender, focus their contribution on the all-important SOA governance issue. They advance a framework identifying five elements of a firm's SOA governance structure: service ownership, service orchestration, service alignment, service delivery, and service value. Armed with the framework, they interpret the Cutter survey results and take stock of the current state of SOA governance. With the survey results in hand, they suggest five critical questions to ask as you approach the design and introduction of the appropriate SOA governance structure for your organization.
Mike's contribution is more general and focuses on identifying best practices in SOA adoption and use. He identifies the expectations and challenges associated with SOA projects and then advances a set of six areas in which best practices can be identified: reference architecture, common semantics, governance, business process modeling, design time repositories, and model-based development. Mike intersperses his piece with interesting sidebars that I particularly draw to your attention.
