Energy Efficient VM Placement Heuristic Algorithms Comparison for Cloud with Multidimensional Resources

摘要

Cloud computing provides user utility-oriented IT services, yet accompanied with huge energy consuming, which contributes to the high operational cost as well as CO_2 emission. Making Cloud computing energy efficient can lead to a better tradeoff between profit and environmental impact. In this paper, we formulate the energy efficient VM placement problem in Cloud architecture with multidimensional resources and introduce the objective of this problem. Heuristic algorithms including three traditional local search algorithms and generic algorithm (GA) are presented to provide possible optimized solution. We conduct experiments based on Cloudsim. The result shows that GA sometimes provide the best solution, but with poor stabability. Although the BF provide neither the best nor the worst solution most of time, it have the best stabability.