Energy Efficient Storage Management Cooperated with Data Intensive Applications in Virtual Machines

In data centers, huge amount of computes are running. These computes consume enormous energy. For this issue, an energy efficient storage management method cooperated with data intensive applications was proposed. With this method, data and storage devices are managed with application supports and power consumption of storage devices is significantly decreased. However, the work does not take account of virtualized environment. Recently, many data intensive applications run on virtualized environment, such as cloud computing environment. Thus, we think discussion on power effective storage management in virtualized environment is also important in addition to this work. In this paper, we focus on virtualized environment in which plural virtual machines run on a physical computer and a data intensive application runs on each virtual machine. We apply this storage management method for this environment, and evaluate performance and power consumption. Then, we proposed two storage placing methods, which are symmetric placing and asymmetric placing, for virtualized environment. Our experiments demonstrated that the proposed methods could create long HDD access intervals enough to save power consumption of storage devices. In addition, our results showed that asymmetric method could better performance than symmetric method.