Abstract
Data centers are constantly growing and evolving in number and complexity due to increasing demand for Internet-based services. As a result, energy consumption in data centers has increased significantly in recent years, which has become a critical concern for IT enterprises and governments because of high operational costs and negative environmental impact. Therefore, green solutions are needed to integrate the use of renewable energy with the development of reduction strategies in energy consumption. In this study, we investigated the performance of a system that can simulate a data center constrained by service-level agreements, energy consumption, power generation, and non-exponentially distributed service times. A discrete event simulation, an optimization model, and a forecasting method were integrated into the system's architecture to analyze performance when facing different scenarios with several changes in the system's characteristics. We conducted a survey on energy trading, considering that renewable energy generators were incorporated into the algorithm to determine the interaction between data centers and smart grids. The experimental results demonstrate that the proposed system has great potential in improving energy efficiency under different operating conditions in the data centers.