Energy efficiency has become a major concern in modern data centers design due to financial and environmental considerations. In existing large data centers, dynamic virtual machine (VM) provisioning between these servers is the most efficient method to improve servers’ utilization and reduce energy. These methods use monitoring, historical data, prediction, and virtual machine placement algorithms to achieve their goals. This paper proposed a real-time framework to optimize the process of VM consolidation base energy consumption. It introduces a novel index named Machine Condition Index (MCI) of a server to measure the extent by which physical machine (PM) is suitable to handle the new (consolidated) VM in large scale heterogeneous data centers. This index represents a group of of weighted components including data center network, physical machines, storage, power system and facilities, used in any cloud data center. It is an easily used tool for comparing server energy consumption and can lead to increased efficiency and manageability of data center resources. © 2019 IEEE.

View