A Multi-start Local Search Scheduler for an Energy-aware Cloud Manager

Yacine Kessaci (1) , Melab Nouredine (2) , El-Ghazali Talbi (1)

1 DOLPHIN - Parallel Cooperative Multi-criteria Optimization
2 LIFL - Laboratoire d'Informatique Fondamentale de Lille

The field of cloud computing uses different management techniques for data center virtualization such as OpenNebula. However, computers composing the cloud infrastructure use a significant and growing portion of energy in the world specifically when dealing with virtualization for high performance computing (HPC). Therefore, energy-aware computing is crucial for large-scale systems that consume considerable amount of energy. In this paper, we present a new work that aims to deal with the energy consumption within a realistic cloud infrastructure using OpenNebula as a software management solution. Our scheduler is based on a multi-start local search heuristic that helps to find the best scheduling by dispatching the arriving of virtual machines (VM) according to the minimum energy consumption.

Domaines: Calcul parallèle, distribué et partagé [cs.DC]

Yacine Kessaci, Melab Nouredine, El-Ghazali Talbi. A Multi-start Local Search Scheduler for an Energy-aware Cloud Manager. META 2012, Oct 2012, Sousse, Tunisia. ⟨hal-00749063⟩