Environmental Impact Assessment And Optimization Strategies for Energy-Efficient Cloud Data Centers

Abstract

The increasing number of cloud services has worsened the environmental footprint of large-scale data centers, especially in the energy consumption and large carbon footprint. In this work, we introduce a multi-dimensional green aware optimization framework, involving the integration of dynamic virtual machine (VM) consolidation based on virtualization, SLA-aware scheduling and renewable energy incorporation into an otherwise un-integrated multi-dimensional optimization, in order to maximize the energy efficiency and the green-ness of cloud data center operations. The model is simulated and tested in CloudSim by using real-world users workloads tracings in PlanetLab and renewable generation characterization in National Renewable Energy Laboratory. Power and emission factor models are estimated to assess energy consumption and carbon emissions, and performance is measured based on energy savings, CO 2 reduction, status of log violations, use of renewable energy sources and computational overhead. Findings confirm that a suggested system decreases total energy demand and greenhouse gas emissions by 28.2 percent and 41 percent, respectively, relative to the baseline operations with renewable contribution to the total demand of over 35 percent. Results indicate the combined optimization of resource distribution, service provision, and energy supply provide better results in sustainable performance than the conventional myopic approach.