Energy efficient cloud data center management based on fuzzy multi criteria decision making

Abstract

In computing cloud the ultimate goal is to get optimal performance by utilizing minimum computing resources, this can be achieved by avoiding wasting of resources as a result of under-utilization and to cut lengthy response time, due to over-utilization. Power consumption among cloud data centers is growing at a rapid pace and efficient energy management is one of the most challenging research issues. Furthermore, Virtualization and live migration of VMs among PMs Plays a vital role in data center load management. In this paper we propose and evaluate an approach for power and performance management through intelligent decisions over hotspot detection and migration time of VMs across heterogeneous PMs for mitigation of hotspot of a cloud computing data center. We used two dynamic threshold levels (peak and off-peak) load strategy to implement our decisions. The focus of this paper is to present a new method to find overloaded nodes at upper level threshold (peak load), and also perform PMs consolidation at lower-level threshold and putting idle server to sleep state, by using Best Fit Decreasing based on TOPSISone of the most efficient Multi Criteria Decision Making techniques. Finally, we perform simulation to evaluate the work and results show that proposed method brings substantial energy saving, while ensuring reliable QoS. © 2013 Springer-Verlag.