Elastic Power Utilization in Sustainable Micro Cloud Data Centers

Abstract

Efficient utilization of renewable energy when powering Cloud Data Centers is a challenging problem due to the variable and intermittent nature of both workload demand and renewable energy supply. This work aims to develop an innovative dynamic resource management algorithm to provide energy flexibility to data center operators for shaping their energy demand to match renewable energy supply. We present a novel framework, called Elastic Power Utilization ( EPU ), to serve this purpose. EPU utilizes energy source information to dynamically manage data center resources for matching the renewable energy supply with the energy demand to serve the workload. We propose a resource management algorithm that exploits overbooking, consolidation and migration of virtual machines (VMs) to implement the power elasticity required by the EPU framework. We compare our approach to a state-of-the-art algorithm and baseline approaches with three different workloads. The results from extensive simulations show that our proposed algorithm outperforms the state-of-the-art approach in saving brown energy by 23.1%, 21.3%, and 27.0% for Google , Wikipedia , and Nectar workloads, respectively.