Energy-aware vm migration in cloud computing

Abstract

The continuous growth of cloud data centers is accompanied by enormous amounts of energy consumption leading to carbon dioxide (CO2) emissions making the environment unfriendly. Dynamic VM consolidation is an effective policy that can reduce energy consumption in data centers. It reduces the number of servers used complying with Quality of Service (QoS) constraints. This paper presents two-phase approach: First, Power-Aware Placement of VMs provides the least increase in power consumption. Second, a function of CPU utilization and memory utilization with double-threshold policy is used to estimate host current utilization. It chooses VMs for migration based on their CPU and memory utilization, thus reducing the chances of SLA violation and number of migrations as minimum as possible. The energy consumption is optimized by controlling the resource utilization and shifting the idle servers to sleep state. Simulation results depict that our proposed approach significantly reduces energy consumption in dynamic workload scenarios when compared with other algorithms.