An Efficient Task Scheduling Strategy for DAG in Cloud Computing Environment

Abstract

Cloud computing is an active research topic in computer science and its popularity is increasing day-to-day due to the high demand of cloud in every field. Data center in cloud platform is having the number of computing resources which are interconnected with very high-speed network. These resources are accessed at the rapid speed so that minimum interaction with service provider. Task scheduling is a burning area of research in cloud environment. Here an application program is represented by directed acyclic graph (DAG). Major concerned of the task scheduling method is to reduce overall execution time. i.e., to minimize the makespan. This paper presents a new strategy for task scheduling in DAG which based on two well-known attributes critical path and static level. By using these attributes, we have developed new attributes CPS which is summation of critical path and static level. New strategy works on two phases such as task priority and resource selection. The proposed method is tested using two DAG models which shows outperformance as compared to heuristic algorithm HEFT. Comparisons have been done using some performance metrics which also gives good result of proposed method.