Opportunistic energy-aware rescheduling in desktop grid environments

Nowadays either maximizing energy efficiency and improving resource utilization is a challenge among the different existing distributed systems, specially in large scale distributed environments such as Grids or Clouds. With this objective, we propose a rescheduling technique that tries to improve resource usage, whilst at the same time tries to minimize the energy needed for the executions of the already accepted jobs by using first/more the resources that are more energy efficient and without reducing the QoS provided. The information obtained from a device capable of measuring the energy that each desktop resource needs is used by the algorithm at the resource selection process, resulting in a noticeable reduction in the energy used as it has been demonstrated in a real desktop Grid environment.