Cloud Workload Prediction by Means of Simulations

Clouds hide the complexity of maintaining a physical infrastructure with a disadvantage: they also hide their internal workings. Should users need to know about these details e.g., to increase the reliability or performance of their applications, they would need to detect slight behavioural changes in the underlying system. Existing solutions for such purposes offer limited capabilities. This paper proposes a technique for predicting background workload by means of simulations that are providing knowledge of the underlying clouds to support activities like cloud orchestration or workflow enactment. We propose these predictions to select more suitable execution environments for scientific workflows. We validate the proposed prediction approach with a biochemical application.