Model-driven synthesis of SOA solutions

  • Authors:
  • J. K. Strosnider;P. Nandi;S. Kumaran;S. Ghosh;A. Arsanjani

  • Affiliations:
  • IBM Integrated Supply Chain (ISC), Chapel Hill, NC;IBM Thomas J. Watson Research Center, Yorktown Heights, NY;IBM Software Group, Somers, NY;IBM Global Business Services, Tampa, FL;IBM Global Business Services, Cedar Rapids, IA

  • Venue:
  • IBM Systems Journal
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

The current approach to the design, maintenance, and governance of service-oriented architecture (SOA) solutions has focused primarily on flow-driven assembly and orchestration of reusable service components. The practical application of this approach in creating industry solutions has been limited, because flow-driven assembly and orchestration models are too rigid and static to accommodate complex, real-world business processes. Furthermore, the approach assumes a rich, easily configured library of reusable service components when in fact the development, maintenance, and governance of these libraries is difficult. An alternative approach pioneered by the IBM Research Division, model-driven business transformation (MDBT), uses a model-driven software synthesis technology to automatically generate production-quality business service components from high-level business process models. In this paper, we present the business entity life cycle analysis (BELA) technique for MDBT-based SOA solution realization and its integration into service-oriented modeling and architecture (SOMA), the end-to-end method from IBM for SOA application and solution development. BELA shifts the process-modeling paradigm from one that is centered on activities to one that is centered on entities. BELA teams process subject-matter experts with IT and data architects to identify and specify business entities and decompose business processes. Supporting synthesis tools then automatically generate the interacting business entity service components and their associated data stores and service interface definitions. We use a large-scale project as an example demonstrating the benefits of this innovation, which include an estimated 40 percent project cost reduction and an estimated 20 percent reduction in cycle time when compared with conventional SOA approaches.