Macehiter Ward-Dutton –
|Setting the stage for SOA|
|SOA - why now?|
|About the authors|
|About Macehiter Ward-Dutton|
|Download the complete report|
This analyst report from Macehiter Ward-Dutton explores the various perspectives tied up with the question of what constitutes an IT service and shows how to avoid the risks of missing the real point of SOA. The sections excerpted here discuss how SOA can be used to improve IT-business alignment, tips for a successful SOA initiative, and the role of Web services. Setting the stage for SOA
The common understanding of SOA today (a particular arrangement of middleware technology which provides infrastructure for integrating and composing Web services) is a simple one. Pursuing an IT initiative based on this understanding can bring benefits. But SOA in this context will not by itself deliver significant improvements in the alignment of IT with business needs. The concept is too limited in scope and too focused on software development issues.
Improving "IT-business alignment" requires an evolution both in technology and technology thinking, which makes it easier for organizations to more effectively exploit their existing IT assets and re-prioritize IT investment, delivery and strategy so that IT assets can be more readily directed to deal with business change.
This evolution depends on a service-based approach to the delivery of IT, but the current focus of SOA "innovation" is far too narrow. The danger of the "simple view of SOA" is one of over-inflated expectations. Business and IT professionals alike are being targeted by vendors which position SOA as "something that you can buy" that can make a significant positive impact on the alignment of IT and business. But in fact, a SOA initiative which addresses a discrete IT project, if considered in isolation, will most likely fail to deliver the kinds of benefits you might expect, and lead quickly to disillusionment. SOA is an evolved approach to thinking about how IT services are designed and delivered: and the business case for adopting it should be made in this context.
In order for an SOA initiative to really improve IT-business alignment, it has to be considered as part of a larger picture, which encompasses and links different perspectives of IT services. The truth is that different groups of people have different ideas of what an IT service is, and, most importantly, IT users' perspectives of IT services are very different from those of software developers considering SOA.
A true picture of "service-oriented IT," which puts managed IT services at the heart of the relationship between business and IT, requires you to think about all aspects of enterprise IT provision from a "service-oriented perspective" - not just those software-based services that are provided by business application functionality, but those services that are provided by people, too.
However, the vast majority of today's SOA initiatives focus firmly on one type of IT service: business function services (software services which automate aspects of particular business functions). But there are two other types of IT service which need to be considered: infrastructure services (which provide the underlying platform over which business function services are delivered, and which tend to be the focus of IT operations professionals) and lifecycle services (which are responsible for the design, implementation, operation and alteration of infrastructure and business function services, and which tend to be the focus of business investment in IT).
Anyone ready to embark on an SOA initiative needs to think about where SOA can take them, as well as worrying about how to get the initiative off the ground. This means engaging suppliers which can demonstrate insight not only into the software development issues; but also into the operational issues, and into the commercial and business issues that can arise when implementing online services on a large scale.
Implementing an SOA initiative which is going to have longevity, and which is going to improve IT-business alignment, means thinking about the complications which come with large-scale deployment, earlier rather than later. In this report we identify three particular issues which need to be considered. Firstly, the design of business function services using knowledge of business processes as the driver; secondly, the adoption of a policy-based approach to operational assurance, based on contracts; and thirdly, the use of a variety of approaches to align service quality to the roles of individual services.
Service-oriented architecture (SOA) - why now?
A long history
After a considerable period, thankfully, more and more industry influencers and commentators are recognizing that the basic ideas encompassed in service-oriented architecture (SOA), which has generated huge interest over the past couple of years, are not at all new. These ideas have roots stretching back at least thirty years. David Parnas' seminal 1972 paper "On the Criteria to Be Used in Decomposing Systems Into Modules"; object-oriented programming languages Eiffel, C++, java and Smalltalk; the Distributed Computing Environment (DCE) standard; Microsoft's COM and the ill-fated OpenDoc initiative; Microsoft's DCOM and the OMG's Common Object Request Broker Architecture (CORBA standard); distributed computing middleware platforms like Tuxedo, Encina, and TopEnd; all these refer to, or rely on, key ideas that are commonly expressed in today's notion of SOA.
Web Services brings new color to the picture
So if all this is "old hat", why has interest been re-ignited? From technology suppliers' point of view, the answer, broadly, is the availability of a set of application-level communication protocols (SOAP, WSDL, UDDI and others) collectively known as Web Services technology. SOA hasn't purely been driven by suppliers' use of and interest in Web Services, however: SOA has also been helped along by industry interest in any cost-effective technology that might help to ease enterprises' application development and integration burdens.
Web services technology has added new color to the SOA picture, because it has the potential to enable service-oriented software integration and development that act at multiple different levels of scope and abstraction independently of the underlying implementation technology. Whereas most of the earlier middleware technologies provided benefits at the levels of individual complied programs or possibly distributed systems, Web services technology has the potential to glue software together across industry, organization and geographical boundaries as well as across application, system and program boundaries.
The support that Web services technology has garnered across the IT industry has helped to fuel its universality further. Quite rapidly, IT suppliers of many types have started to use the technology to reshape their products - and the key result of this is that a broad range of historically distinct IT domains are now starting to become linked together around the common concept of "service." Particularly:
- Application development technology is starting to focus on the creation of "composite applications" which derive much of their functionality from facilities provided by existing Web services
- Application integration technology is starting to focus on connectivity and orchestration of resources which are exposed using Web services
- Systems management technology is starting to focus on providing information and insight into operational IT behavior at an increasingly high-level with a focus on managing "service levels" - which is variously branded as service level management (SLM), IT service management (ITSM), and business service management (BSM).
- Information management technology is starting to focus on delivering access to information (both structured and unstructured) and associated functionality exposed as Web services.
It is this "convergence" of IT domains, driven by common use of Web services technology - rather than just the simple availability of Web services - which is the real key to understanding the potential value of SOA. SOA "done right" is about using service-orientation to pull multiple different IT design, delivery, operations and change management disciplines together within an IT organization to serve business needs.
About the authors
Neil Macehiter's areas of expertise include enterprise architecture, service-oriented architecture (SOA), Web services, virtualization and identity management. He has acted as an advisor to leading vendors, including IBM, Oracle, Microsoft, and Sun Microsystems; and to large IT user organizations in Government and the private sector.
Neil Ward-Dutton's areas of expertise include application development, business integration tools, process management and application platforms, enterprise architecture and service oriented architecture (SOA), has acted as an advisor to leading vendors, including IBM, Oracle, Microsoft, BEA, Hewlett-Packard, SAP, and Borland; and to large IT user organizations in Media, Government, Transport, Financial Services, and Telecommunications.
About Macehiter Ward-Dutton
Macehiter Ward-Dutton is a specialist IT advisory firm that focuses exclusively on issues concerning IT-business alignment. We use our significant industry experience, acknowledged expertise, and a flexible approach to advise businesses on IT architecture, integration, management, organization, and culture.