KWH-PRO: um arcabouço para alocação dinâmica de vms e redução do consumo de energia

Abstract

Currently, there is a growing need for data storage and processing. This has been driving the roll-out of data centers around the world. These centers are estimated to account for about 2% of the world’s electricity consumption, and the trend is that consumption will continue to increase. Thus, the need for research in this area is evident. This work presents a framework that has as main goal the reduction of energy consumption in data centers with VMs and that the economy interferes minimally in the performance of these centers. In order to achieve the objective, the framework automatically performs the following tasks: (i) monitoring of the energy consumption and workload of servers, (ii) classification of servers using the monitoring data, (iii) VMs with hot migration and (iv) server bindings and shutdowns. To validate the scaffold efficiency, a computational cloud implemented with OpenStack was used. Several experiments were performed, in which energy savings of up to 38% could be observed, at the cost of a slight degradation of cloud performance that seemingly imperceptible.