Extending Energetic Potentials of Data Centers by Resource Optimization to Improve Carbon Footprint

The electric power is one of the major operating expenses in data centers. Rising and varying energy costs induce the need of further solutions to use energy efficiently. The first steps to improve efficiency have already been accomplished by applying virtualization technologies. However, a practical approach for data center power control mechanisms is still missing. In this paper, we address the problem of energy efficiency in data centers. Efficient and scalable power usage for data centers is needed. We present different approaches to improve efficiency and carbon footprint as background information. We propose an in-progress idea to extend the possibilities of power control in data centers and to improve efficiency. Our approach is based on virtualization technologies and live-migration to improve resource utilization by comparing different effects on virtual machine permutation on physical servers. It delivers an efficiency-aware VM placement by assessing different virtual machine permutation. In our approach, the applications are untouched and the technology is non-invasive regarding the applications. This is a crucial requirement in the context of Infrastructure-as-a-Service (IaaS) environments.