A Power-Aware Scheduling of MapReduce Applications in the Cloud

Ying Li,Hongli Zhang,Kyong Hoon Kim

DOI: https://doi.org/10.1109/DASC.2011.111

2011-01-01

Abstract:Cloud computing is an emerging computing technology for large data center that maintains computational resources through the internet, rather than on local computers. The large data centers maintain Cloud computing applications with lots of cost because of power consumption, which results in a new research issue, called Green Cloud computing. Since MapReduce is one of popular Cloud computing models, this paper focuses on how to reduce energy of MapReduce applications. Thus, we propose a new power-aware MapReduce application model to be used for power-aware computing with consideration of users' requirements. We also provide a scheduling algorithm for MapReduce applications in heterogeneous Cloud resources and suggest power-aware schemes in order to reduce the total energy. Throughout simulation results, we show that the proposed scheduling algorithm saves more energy than static schemes.