Delayed Best-Fit Task Scheduling to Reduce Energy Consumption in Cloud Data Centers

Reducing energy consumption of cloud data center is critical for its sustainable growth. We propose the delayed best-fit task-scheduling scheme that strategically delays the scheduling of tasks to the most energy-efficient servers of data centers to reduce its energy consumption. The proposed scheme uses static and dynamic thresholds mechanisms to an allocated task to an assigned server to balance energy consumption and task completion time. The proposed scheme is tested on a real traffic trace from a Google data center and compared with best-fit and first-fit scheduling algorithms. We show that the proposed delayed best-fit task-scheduling scheme reduces data center energy consumption by 15% of that attained by the best-fit algorithm on the same trace, without compromising the average task completion time.