Skip to Main Content
When various technical and business changes related to Web service compositions occur, there is often a need for runtime adaptation with minimal human intervention. While many research projects work on particular types of such adaptation, much more research is needed on decision making for diverse autonomic adaptations, particularly to maximize business (as opposed to technical) metrics. Our MiniMASC middleware is a framework for diverse autonomic adaptations of Web service compositions, focusing on supporting such advanced decision-making algorithms. MiniMASC is relatively simple, light weight, modular, and extensible. It uses the WS-Policy4MASC policy language that can describe all information necessary for different types of adaptation. After presenting our classification of different types of decision making in adaptation of Web service compositions, this paper discusses how MiniMASC (with WS-Policy4MASC) can be used for implementing algorithms maximizing business metrics. Our tests show that the implementation of MiniMASC has a satisfactory performance and scales well.