Energy Efficient and Performance Aware Multi-Objective Allocation Strategy in Cloud Environment

In Cloud Computing, high Energy Consumption and Service Level Agreement violations are the challenging issues that need to be addressed. Efficient resource management strategies are required to utilize resources in a better way. There are number of approaches focusing either on issue of high Energy Consumption or Service Level Agreement violations. But, the trade-off between the two is highly ignored. In this work, Energy and Performance Aware Best Fit Allocation policy is suggested that attempts to accommodate workload on already active energy efficient servers, keeps the idle servers to low power mode and, ensures minimum free leftover MIPS on the server to utilize it properly. Moreover, it attempts to keep some unallocated MIPS on the server to make the server capable to adjust to some sudden increase in resource requirements by the applications thus minimizing the need of frequent VM reallocation that in turn contributes to lower Service Level Agreement violations that are caused due to migrations. The experimental results have shown an average reduction of 24.62% and 63.37% in energy consumption and migration count respectively.