Cloud workload prediction and generation models

Gilles Madi Wamba (1) , Yunbo Li (1, 2) , Anne-Cécile Orgerie (3, 2) , Nicolas Beldiceanu (1, 4) , Jean-Marc Menaud (5, 1)

1 IMT Atlantique - DAPI - Département Automatique, Productique et Informatique
2 MYRIADS - Design and Implementation of Autonomous Distributed Systems
3 CNRS - Centre National de la Recherche Scientifique
4 LS2N - équipe TASC - Théorie, Algorithmes et Systèmes en Contraintes
5 LS2N - équipe ASCOLA - Aspect and Composition Languages

Cloud computing allows for elasticity as users can dynamically benefit from new virtual resources when their workload increases. Such a feature requires highly reactive resource provisioning mechanisms. In this paper, we propose two new workload prediction models, based on constraint programming and neural networks, that can be used for dynamic resource provisioning in Cloud environments. We also present two workload trace generators that can help to extend an experimental dataset in order to test more widely resource optimization heuristics. Our models are validated using real traces from a small Cloud provider. Both approaches are shown to be complimentary as neural networks give better prediction results, while constraint programming is more suitable for trace generation.

Mots clés: workload; constraint programming; generation; prediction; machine learning; artificial neural network; cloud; green energy

Domaines: Informatique [cs]

Gilles Madi Wamba, Yunbo Li, Anne-Cécile Orgerie, Nicolas Beldiceanu, Jean-Marc Menaud. Cloud workload prediction and generation models. SBAC-PAD 2017 : 29th International Symposium on Computer Architecture and High Performance Computing, Oct 2017, Campinas, Brazil. pp.89-96, ⟨10.1109/SBAC-PAD.2017.19⟩. ⟨hal-01578354⟩