Design and Evaluation of an Autonomic Workflow Engine

In this paper we present the design and evaluate the performance of an autonomic workflow execution engine. Although there exist many distributed workflow engines, in practice, it remains a difficult problem to deploy such systems in an optimal configuration. Furthermore, when facing an unpredictable workload with high variability, manual reconfiguration is not an option. Thanks to its autonomic controller, the engine features self-configuration, self-tuning and self-healing properties. The engine runs on a cluster of computers using a tuple space to coordinate its various components. Its autonomic controller monitors its performance and responds to workload variations by altering the configuration. In case failures occur, the controller can recover the workflow execution state from persistent storage and migrate it to a different node of the cluster. Such interventions are carried out without any human supervision. As part of the results of our performance evaluation, we compare different autonomic control strategies and discuss how they can automatically tune the system