A service is designed to interact with other services. If the service interaction is stateful and asynchronous, the interaction protocol can become quite complex. A service may be able to interact with a lot of possible partner services, one partner or no partner at all. Having no partner surely is not intended by the designer. But the stateful interaction between services can be formalized and thus analyzed at design time. We present a formalization which is centered around a graph data structure that we call interaction graph, which represents feasible runs of a partner service according to the interaction protocol. As interaction graphs suffer from state explosion, we introduce a set of suitable reduction rules to alleviate the complexity of our approach. As our case studies show we are able to analyze the interaction behavior of a service efficiently.
CITATION STYLE
Weinberg, D. (2009). Efficient controllability analysis of open nets. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5387, pp. 73–87). https://doi.org/10.1007/978-3-642-01364-5_5
Mendeley helps you to discover research relevant for your work.