Comparison of composition engines and identification of shortcomings with respect to cloud computing

Abstract

Most workflow engines are currently not Cloud-aware. This is due to multiple reasons like no support for transparent scalability, no multi-tenancy support, no ability to store process related data in a Cloud storage, or no support for quality of service enforcements. Recently Cloud based workflow services appeared in the workflow landscape and promise to run workflows in the Cloud. This student reports evaluates current state of the art BPEL and BPMN workflow engines and Cloud based workflow services according to their Cloud- awareness and general workflow functionalities. Identified shortcomings are described and prioritized. As a result of this evaluation the workflow engine WSO2 Stratos is best suited for running workflows in the Cloud, but it lacks native clustering support and quality of service enforcement.