Queuing-based capacity planning approach for Web service workflows using optimization algorithms

One major challenge for service-oriented workflows in digital ecosystems is capacity planning for cross-organizational Web service workflows in order to avoid performance degradation. In order to analyze the execution capacity of Web service workflows and to plan the workflow control, queuing theory can be used to describe the average performance behavior of a workflow. By addressing capacity planning of Web service workflows, resource usage becomes more and more important. Capacity planning and performance measurement are crucial to ensure that the workflow execution remains feasible and SLA violations due to overload are avoided. Thus, this paper presents a capacity planning approach for Web service workflows based on queuing theory to support capacity planning decisions. Further we describe an optimization algorithm how to achieve an optimal utilization of the invoked Web services at minimal costs.