Energy reducing dynamic multi-dimensional resource allocation in cloud data center

Abstract

Consolidation virtual machine (VM) on fewer physical servers is an efficient way to reduce energy consumption in data center. However, VMs often have inherent dependencies which bring complex load interactions among the underlying physical servers. Unreasonable migration will lead to a decline in network performance. In this paper, aiming at energy saving, we propose an application-aware policy in VM consolidation process. Moreover, in order to guarantee that the migrated VMs can be efficiently placed on physical servers determined by the application-aware policy, we also propose a multi-dimensional resource allocation algorithm (MDRA). Experiment results show that the proposed application-aware policy can efficiently reduce energy consumption as well as improving network performance. And the proposed MDRA algorithm can ensure the effectiveness of our policy.