Green Data Centers through Power Management of Underutilized and Idle Resources

Abstract

Cloud computing proved to be that promising technology which enabled the world to bring resource sharing to a new level by turning the internet into a cloud. With the rise of Internet-scale systems and cloud computing services, there is an increasing trend towards building massive, energy-hungry, and geographically distributed data centers. Due to their enormous energy consumption, data centers are expected to have major impact on the electric grid and potentially the amount of greenhouse gas emissions and carbon footprint. The aim of this research is to make data centers the heart and core of cloud computing energy efficient and green. For this purpose the main focus is on achieving this through the management of idle and underutilized resources. Virtualization is used for making power aware and green usage of the resources. In an attempt to do so a Wholesome Green Approach is presented and analyzed. This approach uses a new scheduler and a modified VM allocator combination to achieve better allocation of load to the servers. Green Scheduler first balances the resource request load to the data center among the available VMs in a way that all are fully utilized and the first ones are filled first. Later the VMs are managed through Green VM Allocator which checks how many servers are running how many VMs and if they can be consolidated to free resources or to maximize utilization of the available servers. The end result is clean data center where the powered on resources are fully utilized and the idle resources are identified and turned off to lower the energy consumption and carbon footprint of the data center.