An Effective Energy and Performance Management of Cloud Data Canters using Virtual Machine Consolidation

Abstract

Establishing large-scale data center increases demand in computational power of cloud computing. The data centers consume a lot of electricity for continuous customer service support. Data center electricity consumption is projected to increase to roughly 140 billion kilowatt-hours annually by 2020, the equivalent annual output newlineof 50 power plants, costing American businesses $13 billion annually in electricity newlinebills. So cloud computing datacenters extensively applied a virtualization technology newlinewhich is realizing energy efficient operations of cloud datacenter. Virtual machine newline(VM) consolidation technology is widely applied to modern data center for cloud newlinecomputing as a key technology to realize energy-efficient operation of a datacenter newlineserver. Consolidation achieves energy efficiency by enabling multiple instantiations newlineof operating systems (OSs) to run simultaneously on a single physical machine. newlineWhile, live migration of virtual machine can transfer the virtual machine workload newlinefrom one physical machine (PM) to another without interrupting service. However, newlineboth the technologies have their own performance overheads. Here there is a need of newlinegood tradeoff between the performance and energy efficiency. In this work, we study newlinethe energy efficiency from the performance perspective. There are three contributions newlinein the frame work. First, PM overload detection based on Service Level Agreement newlineViolation (SLAV) and under load detection based on minimum number of VM for newlineVM allocation algorithm is established. Second, in VM selection algorithm VM with newlinemaximum utilization for migration has been selected. Third, VM placement algorithm newlineusing Minimum Power High Available Capacity algorithm for finding new placement newlineof VM has been developed. We have evaluated our frame work in Cloud Sim newlinesimulation tool. The potential performance metrics caused by VM consolidation and newlinemigration of virtual machine technology has also been investigated. Experimental newlineresults show that both the technologies can effectively implement energy-saving goals newlinewith little performance overheads. An effective VM consolidation and migration newlinestrategies can improve the energy efficiency. newline