Several companies have invested vast amounts of resources and effort into information systems. In the course of time, these systems have integrated with idiosyncratic point-to-point solutions to address larger business transactions. While this has appeared as simple and pragmatic on the outset, over time the increasing complexity and disconnected dependencies have resulted in “integration spaghetti” that is prohibitively difficult and operationally expensive to manage and maintain. The problem is aggravated by the increasing need for business agility that cannot be achieved with cast-in-concrete proprietary solutions.
While this deeply ingrained “technology mess” is too overwhelming to be dismantled and replaced, the truth is out there that and can be disentangled. By mapping physical data to logical information objects and technical functions to business activities, a canonical information and operational model can be elicited. To this end, a business has to define its own, consistent ontology, which may partly be adopted from the normative nomenclature of this respective industry vertical (Health care, Financial and Insurance, etc..). The data in the underlying information systems (both silo’ed and horizontal), can be mapped to these canonical concepts using appropriate integration technologies and exposed as data or process centric SOA services. By analyzing the functional requirements of business processes, a set of logic centric services can also be identified.
In Service Oriented Architecture (SOA), the intricacies of applications and technology infrastructure are encapsulated behind well defined, self-describing service interfaces that exposed the contained information and functionality as reusable, context independent services. The underlying implementation of a service can be changed as long as the service contract is maintained. SOA services provide modular building blocks that can be composed to higher level constructs, e.g., orchestrations or composite applications. The coarse-grained modularity reduces inherent complexity of the enterprise architecture and improves business agility. The service abstraction insulates business changes from IT development and thereby bring more synergies between the business and IT life-cycles.
I will be writing on how to enable an agile SOA with BPM and Event Driven architectures in the next segment. Good day.
Filed under: SOA
