Datasheet

P1: OTA/XYZ P2: ABC
c01 JWBT191/Bell November 1, 2009 14:39 Printer Name: Yet to Come
Service-Oriented Discovery and Analysis Modeling 15
PATTERNS OF IMPLEMENTATION: BEST PRACTICES AND OUT-OF-THE-BOX SOLUTIONS. Dur-
ing service-oriented discovery and analysis, employ the provided patterns for service identification
and inspection. These are the best practices and predefined solutions that can reduce development
cost, analysis and discovery time, and ultimately organizational expenditure. Use these repeatable
templates to:
r
Shape the solutions for a project or a business initiative
r
Govern the overall service discovery and analysis process
r
Model a virtual service environment, mimic service functionality, and simulate service
integration
These patterns of design and implementation are vital to service-oriented discovery and anal-
ysis because they reduce implementation risks and introduce road-tested, out-of-the-box solutions
that have been applied in similar circumstances. They are the guiding pillars of the analysis pro-
cess and major contributors to shaping a service-oriented analysis proposition. An overview of the
service-oriented discovery and analysis patterns is provided later in the Service-Oriented Discovery
and Analysis Patterns section later in the chapter.
SERVICE-ORIENTED DISCOVERY AND ANALYSIS MODELING
An analysis proposition is articulated by an analysis proposition modeling diagram that depicts vital
design perspectives of a service and its hosting environment. These views include information about
internal service composition, such as aggregated components, service operations and interfaces,
service internal boundaries, service specialties, service capabilities, and more. Service ecosystem
external aspects, such as deployment, service coupling, contract model, distribution model, and
interoperability mechanisms should also be illustrated in an analysis proposition diagram.
Furthermore, a proposed analysis solution must be presented in a formal manner, in which
the design of services and their corresponding distributed environment adheres to a prescribed no-
tation that captures relationships between services and associated consumers. This language also is
devised to describe the evolution of a service and it transformation stages, starting at its inception,
tracing its development phases, and depicting its maturity in production. Chapters 14–22 discuss in
details modeling techniques, notations, and modeling operations.
MODELING LANGUAGE WITH TRANSPARENCY CAPABILITIES. Because analysis aspects dom-
inate this exercise, the language that is employed to present the analysis proposition must trace
decisions that are made during an analysis process. These questions include: Why should a service
be decomposed? Why should a service be retired? Why should two or more services be unified?
Why should a service be further abstracted? What is the driving motivation for decoupling two or
more services? In which circumstances should a binding contract be established between a service
and its corresponding consumers?
Furthermore, as we learned in the previous section, the service-oriented analysis and dis-
covery transparency model also calls for an elaborated analysis proposition that traces business
decisions, technological and architectural best practices and implementations, and even events dur-
ing run-time in production. From a modeling perspective, these may have occurred in the past, be
taking place at the present time, or be planned for the future. This analysis traceability approach is
tuned to the time aspects of a service life cycle. That is, it describes the metamorphosis of a service
since its inception, identifies its existing state, or provides an architecture plan for the future.
Exhibit 1.7 illustrates the three service-oriented discovery and analysis modeling methods,
each of which presents a different state in the service life cycle: to be, as is, and used to be. These
are explained in detail in the sections that follow.