Overcommitment in Cloud Services - Bin Packing with Chance Constraints

40 Pages Posted: 13 Aug 2016 Last revised: 17 Mar 2018

See all articles by Maxime C. Cohen

Maxime C. Cohen

Desautels Faculty of Management, McGill University

Philipp Keller

Google Inc.

Vahab Mirrokni

Google Research

Morteza Zadimoghaddam

Google Inc.

Date Written: August 12, 2016

Abstract

This paper considers a traditional problem of resource allocation: scheduling jobs on machines. One such recent application is cloud computing, where jobs arrive in an online fashion with capacity requirements and need to be immediately scheduled on physical machines in data centers. It is often observed that the requested capacities are not fully utilized, hence offering an opportunity to employ an overcommitment policy, i.e., selling resources beyond capacity. Setting the right overcommitment level can yield a significant cost reduction for the cloud provider, while only inducing a very low risk of violating capacity constraints. We introduce and study a model that quantifies the value of overcommitment by modeling the problem as a bin packing with chance constraints. We then propose an alternative formulation that transforms each chance constraint to a submodular function. We show that our model captures the risk pooling effect and can guide scheduling and overcommitment decisions. We also develop a family of online algorithms that are intuitive, easy to implement, and provide a constant factor guarantee from optimal. Finally, we calibrate our model using realistic workload data and test our approach in a practical setting. Our analysis and experiments illustrate the benefit of overcommitment in cloud services, and suggest a cost reduction of 1.5% to 17% depending on the provider's risk tolerance.

Keywords: Bin packing, Approximation algorithms, Cloud computing, Overcommitment

Suggested Citation

Cohen, Maxime C. and Keller, Philipp and Mirrokni, Vahab and Zadimoghaddam, Morteza, Overcommitment in Cloud Services - Bin Packing with Chance Constraints (August 12, 2016). Available at SSRN: https://ssrn.com/abstract=2822188 or http://dx.doi.org/10.2139/ssrn.2822188

Maxime C. Cohen (Contact Author)

Desautels Faculty of Management, McGill University ( email )

1001 Sherbrooke St. W
Montreal, Quebec H3A 1G5
Canada

Philipp Keller

Google Inc. ( email )

1600 Amphitheatre Parkway
Second Floor
Mountain View, CA 94043
United States

Vahab Mirrokni

Google Research ( email )

Morteza Zadimoghaddam

Google Inc. ( email )

1600 Amphitheatre Parkway
Second Floor
Mountain View, CA 94043
United States

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
291
Abstract Views
1,536
Rank
190,856
PlumX Metrics