Robust approach for host-overload detection based on dynamic safety parameter

Imene El-Taani (1) , Mohand-Cherif Boukala (1) , Samia Bouzefrane (2) , Anissa Imen Amrous (1)

1 USTHB - Université des Sciences et de la Technologie Houari Boumediene = University of Sciences and Technology Houari Boumediene [Alger]
2 CEDRIC - ROC - CEDRIC. Réseaux et Objets Connectés

Host-overloading detection is an important phase in the dynamic Virtual Machines (VMs) consolidation process. Using machine learning to predict the future workload on a host, is a very promising technique to avoid the overload host situation. In this work, we propose a novel approach for overloaded hosts detection, based on neural network and Markov model. The neural network is trained on a workload data set composed of VMs CPU-utilization history. The trained model is then used to predict the future usage for a given Physical Machine(PM), by summing up the predicted utilization of all its VMs. The confidence of this prediction is measured through a dynamic safety parameter, based on Markov model. The obtained results show that our approach outperforms the state of the art algorithms such as: MAD, IQR and LRR.

Mots clés: VM consolidation; host-overload detection; neural network prediction; Markov model; energy-efficiency; SLA violation

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

Imene El-Taani, Mohand-Cherif Boukala, Samia Bouzefrane, Anissa Imen Amrous. Robust approach for host-overload detection based on dynamic safety parameter. IEEE Ficloud (The 9th International Conference on Future Internet of Things and Cloud), Aug 2022, Rome, Italy. ⟨10.1109/FiCloud57274.2022.00044⟩. ⟨hal-03849513⟩