Modeling and Simulating Reconfigurable Networked Service Composites

Composite services are notoriously prone to failure, this is particularly true for long-running, and data-intensive services. Different composition strategies can be employed to make compositions robust. Any service composition strategy does impact performance at the lower network layer and needs to be assessed. Novel approaches are needed to model and evaluate dynamically reconfigurable service composition strategies. We propose an extensible framework to modeling and simulating the behavior of service compositions.