Interference Aware Scheduling of Real Time Tasks in Cloud Environment

Abstract

Cloud environment uses virtualization technology to increase the resource utilization and scalability. Due to the concurrent running of co-allocated tasks with virtual machines on top a physical machine, performance of some tasks suffers significantly because of interference. To improve the quality of service and reduce the service level agreement violation between user and cloud service provider, we presented an interference aware scheduling approach. Interference aware scheduling approach predicts the effect of interference for execution of a task with respect to the resource usage by co-scheduled tasks in a virtualized environment. This prediction model is being used to schedule tasks to achieve better quality of service in term of task guarantee ratio and priority guarantee ratio. We implemented and validated our approach and the results show that our approach performs significantly better than other approaches.