Resource allocation in a Cloud partially powered by renewable energy sources. (Allocation de ressources dans un Cloud partiellement alimenté par des sources d'énergie renouvelable)

Most of the energy-efficient Cloud frameworks proposed in literature do not consider electricity availability and renewable energy in their models. Integrating renewable energy into data centers significantly reduces the traditional energy consumption and carbon footprint of these energy-hungry infrastructures. As renewable energy is intermittent and fluctuates with time-varying, it is usually under-utilized. We address the problem of improving the utilization of renewable energy for a single data center and investigate two approaches: opportunistic scheduling and energy storage. Our results demonstrate that both approaches are able to reduce the brown energy consumption under different configurations. We extend this work to the context of Edge Clouds and Internet of Things on the use case of data stream analysis. We show how to make Edge Clouds greener with on-site renewable energy production combined with energy storage and performance degradation of the users’ applications.