Scheduling techniques in on-demand grid as a service cloud: a review

Abstract

The Infrastructure as a service (IaaS) Cloud is a customer oriented cloud environment that offers user with computing infrastructures on-demand to be used based on the Cloud computing paradigm of pay-per-use. When the IaaS is now utilized to build a traditional Grid network within the cloud environment, it is now called an on-demand Grid as a service (GaaS) Cloud. In the on-demand GaaS Cloud, a user may use hundred of thousand of Grid nodes to implement a job, therefore manual scheduling is not a feasible scheduling solution. The main objective of this review is to study the various concepts and scheduling algorithms used for the on-demand GaaS Cloud in relation to the scheduling parameters used by existing researches. We also survey the Cloud infrastructures, Grid middlewares and the issues addressed by different researchers in the past within this domain of research. Our contribution will thus be of assistance in understanding the key scheduling algorithms and parameters for potential future enhancements in this evolving area of research.