International Journal of Analysis of Task Scheduling Algorithm in Fine-Grained and Course-Grained Dags in Cloud Environment

Abstract. Cloud computing environment has the ability to scale-up or scale-down the amount of resources provisioned to an application to balance the demand. This elastic nature of the cloud is also limited by the ownership and locality. The emergence of Hybrid cloud (combination of private and public cloud) arises as a solution. In a cloud computing environment, task scheduling becomes more complex due its transparency and flexibility. As the beneficiaries of cloud are small and medium enterprises, cost of using the resources, forms the important factor in task scheduling. The user applications are represented by the Directed Acyclic Graph (DAG) which includes the communication cost and computation cost of the various tasks in the graph. Some applications are termed as communication intensive (course-grained) and the rest are computation intensive (fine-grained). The aim of this paper is to apply the Customer Facilitated Cost-based Scheduling algorithm (CFCSC) to both communication intensive and computation intensive DAGs. This paper also records the performance of the CFCSC for better service