Forecasting the Energy Consumption of Cloud Data Centers Based on Container Placement with Ant Colony Optimization and Bin Packing

Abstract

Currently, the energy consumption in the data centers is a major problem due to the increase in the number of servers and services. Physical machines represent huge energy consumers who work continuously to ensure quality in services through virtual instances. The placement of these instances as containers in hosts has an impact on the minimization of energy consumption. Precisely, the guarantee of an efficient placement without wasting material resources of the hosts, makes it possible to minimize the number of active hosts and thus reduce energy consumption. In this article, we will calculate the energy consumed by a data center using CloudSim before and after applying the placement of containers in the hosts by the First Fit Decreasing algorithm (FFD) of Bin Packing and Ant Colony Optimization (ACO). The results show the FFD's superiority over the ACO in minimizing the energy consumed by any type of Cloud system.