Reliability Modeling and Analysis of Service-Oriented Architectures

Service selection and composition are central activities in service- oriented computing, and the prediction of the QoS attributes of a Service- Oriented Architecture plays a key role to appropriately drive these ac- tivities. Software composition driven by QoS criteria (e.g. optimization of performance, maximization of reliability) has been mostly studied in the Component-Based Software Engineering domain, whereas methodological ap- proaches are not well established in the service-oriented area. Indeed, predic- tion methodologies for service-oriented systems should be supported by auto- mated and e±cient tools to remain compliant with the requirement that most of the activities connected with service discovery and composition must be per- formed automatically. Moreover, the adopted implementation should respect the autonomy and independence of each provider of the services we want to include in our analysis. In this chapter we focus on the modeling and analysis of the reliability attribute in Service-Oriented Architectures, with particular emphasis on two aspects of this problem: (i) the mathematical foundations of reliability modeling of a Service-Oriented Architecture as a function of the re- liability characteristics of its basic elements; (ii) the automatization of service composition driven by reliability criteria.