Using Quantile Regression for Reclaiming Unused Cloud Resources while achieving SLA

Abstract : Although Cloud computing techniques have reduced the total cost of ownership thanks to virtualization, the average usage of resources (e.g., CPU, RAM, Network, I/O) remains low. To address such issue, one may sell unused resources. Such a solution requires the Cloud provider to determine the resources available and estimate their future use to provide availability guarantees. This paper proposes a technique that uses machine learning algorithms (Random Forest, Gradient Boosting Decision Tree, and Long Short Term Memory) to forecast 24-hour of available resources at the host level. Our technique relies on the use of quantile regression to provide a flexible trade-off between the potential amount of resources to reclaim and the risk of SLA violations. In addition, several metrics (e.g., CPU, RAM, disk, network) were predicted to provide exhaustive availability guarantees. Our methodology was evaluated by relying on four in production data center traces and our results show that quantile regression is relevant to reclaim unused resources. Our approach may increase the amount of savings up to 20% compared to traditional approaches.
Liste complète des métadonnées

Cited literature [17 references]  Display  Hide  Download

https://hal.inria.fr/hal-01898438
Contributor : Jean-Emile Dartois <>
Submitted on : Thursday, October 18, 2018 - 2:19:37 PM
Last modification on : Wednesday, February 13, 2019 - 9:53:52 AM
Document(s) archivé(s) le : Saturday, January 19, 2019 - 2:26:51 PM

File

Using_Quantile_Regression_for_...
Files produced by the author(s)

Identifiers

Citation

Jean-Emile Dartois, Anas Knefati, Jalil Boukhobza, Olivier Barais. Using Quantile Regression for Reclaiming Unused Cloud Resources while achieving SLA. CloudCom 2018 - 10th IEEE International Conference on Cloud Computing Technology and Science, Dec 2018, Nicosia, Cyprus. IEEE, pp.89-98, 2018, 〈10.1109/CloudCom2018.2018.00030〉. 〈hal-01898438〉

Share

Metrics

Record views

315

Files downloads

124