Energy Efficient Resource Management in Virtualized Cloud Data Centers

High performance computing infrastructures are a direct result of resource-intensive enterprises and scientific applications, and are massive electric power consumers which, in turn, contribute to excessive CO2 emissions. To achieve energy efficiency, we do have energy efficient hardware, Dynamic Voltage and Frequency Scaling (DVFS), and virtualization of computer resources. Virtualization technology develops Virtual Machines (VMs), reducing the amount of hardware being used and improving utilization of resources. Advantages of virtualization include improved fault and performance isolation between applications sharing the same resource, easy movement of VMs from one physical host to another, and support for hardware and software heterogeneity.The recent cloud computing concept leverages virtualization technology with the ability to provide resources on-demand on the pay-as-you-go basis. Our projected heuristics is for dynamic adaption of allocation of VMs in run-time applying live migration according to current utilization of resources and thus minimizing energy consumption. It can handle strict QoS requirements, heterogeneous infrastructure and heterogeneous VMs. CloudSim toolkit was chosen as a simulation framework. One of the heuristics lead to significant reduction of the energy consumption by a Cloud data centre of 83% in comparison with a non-power aware system and 66% in comparison to a system that applies only DVFS technique. Results showed reduction of operational costs, decrease in carbondioxide footprints and overall energy consumption by modern IT infrastructures.

Provided by: University of Melbourne Topic: Virtualization Date Added: Apr 2010 Format: PDF

Find By Topic