Energy-aware service provisioning in volunteers clouds

Abstract

The goal pursued in this study is to optimise the service provisioning and the energy consumed in a special existing cloud system named SlapOS. The SlapOS cloud innovates in considering that the data centre can be composed by dedicated and volunteer machines. We will use the term volunteer clouds for referring to such organisations. Volunteer clouds offer potential advantages for cloud elasticity, storage of big data and the minimisation of the energy consumption. But, we must manage the potential unavailability of volunteer machines. In this paper, we focus on the design of energy-efficient scheduling approaches in such clouds. Our first contribution is to formulate the scheduling challenges in two new computational problems. These problems are NP-hard. Then, we propose an integer linear programming (ILP) formulation and several greedy heuristics. Finally, we evaluate our approaches throughout various simulations of the SlapOS system in a realistic volunteer computing context.