Energy Modeling of Different Virtual Machine Replication Schemes in a Cloud Data Center

Abstract

In cloud computing systems, users' demands are met through deployment of Virtual Machines (VMs) on Physical Machines (PMs). The rapid growth in cloud service demand has led to the establishment of large-scale virtualized data centers. However data centers consume a large amount of energy resulting in high operating costs and contributing to significant greenhouse gas (GHG) emissions. We propose to use replication of VMs for increasing service reliability and performance. This incurs greater energy consumption for the replicas to run on different machines. We evaluate different replication schemes with respect to job completion time by examining the corresponding energy consumption. We derive closed-form expressions of job completion time and energy consumption by reflecting the structure-state process of job execution with different replication schemes. Our numerical result shows the impact of replication schemes on energy consumption by analyzing the job completion times.