Energy-aware scheduling schemes for cloud data centers on Google trace data

Abstract

In this paper, we propose the most efficient server first (MESF) task scheduling scheme to minimize the energy consumed by data-center servers. MESF allocates and schedules tasks to servers according to the energy profile of servers. Energy consumed by data-center servers constitutes the largest portion of the total data-center energy consumption. The proposed MESF scheme uses resource allocation information and server energy profiles to schedule tasks to the servers with the least virtual power consumption (VPC) increment. We tested our proposed scheme on a real-world trace data set from Google clusters, and the simulation results show that the proposed MESF task scheduling scheme outperforms the random-based and least allocated server first schemes on energy savings.