Cool Cloud: A Practical Dynamic Virtual Machine Placement Framework for Energy Aware Data Centers

With the continuing growth of cloud computing services, power consumption has become one of the most challenging issues in data center environments. With the support of today's virtualization technology, the efficiency and flexibility of data center management can be greatly enhanced, creating great energy saving opportunities. However, effective energy aware design is a non-trivial task, considering the size of the data center, the dynamic fluctuation of workloads and the variation of computing resource requests. In this paper, we propose Cool Cloud: a practical solution for managing the mappings of VMs to physical servers. This framework solves the problem of finding the most energy efficient way (least resource wastage and least power consumption) of placing the VMs considering their resource requirements. Experiment result demonstrates our design can effectively improve data center energy efficiency and scales well to large size data centers. Comparing with industry leading product VMware's Distributed Resource Scheduler (DRS), our design offers better performance in both load balancing and power consumption.