ABACUS: An auction-based approach to cloud service differentiation

Zhenjie Zhang, Richard T B Ma, Jianbing Ding, Yin Yang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

11 Citations (Scopus)

Abstract

The emergence of the cloud computing paradigm has greatly enabled innovative service models, such as Platform as a Service (PaaS), and distributed computing frameworks, such as MapReduce. However, most existing cloud systems fail to distinguish users with different preferences, or jobs of different natures. Consequently, they are unable to provide service differentiation, leading to inefficient allocations of cloud resources. Moreover, contentions on the resources exacerbate this inefficiency, when prioritizing crucial jobs is necessary, but impossible. Motivated by this, we propose Abacus, a generic resource management framework addressing this problem. Abacus interacts with users through an auction mechanism, which allows users to specify their priorities using budgets, and job characteristics via utility functions. Based on this information, Abacus computes the optimal allocation and scheduling of resources. Meanwhile, the auction mechanism in Abacus possesses important properties including incentive compatibility (i.e., the users' best strategy is to simply bid their true budgets and job utilities) and monotonicity (i.e., users are motivated to increase their budgets in order to receive better services). In addition, when the user is unclear about her utility function, Abacus automatically learns this function based on statistics of her previous jobs. An extensive set of experiments, running on Hadoop, demonstrate the high performance and other desirable properties of Abacus.

Original languageEnglish
Title of host publicationProceedings of the IEEE International Conference on Cloud Engineering, IC2E 2013
Pages292-301
Number of pages10
DOIs
Publication statusPublished - 2013
Externally publishedYes
Event1st IEEE International Conference on Cloud Engineering, IC2E 2013 - San Francisco, CA, United States
Duration: 25 Mar 201328 Mar 2013

Other

Other1st IEEE International Conference on Cloud Engineering, IC2E 2013
CountryUnited States
CitySan Francisco, CA
Period25/3/1328/3/13

Fingerprint

Distributed computer systems
Cloud computing
Scheduling
Statistics
Experiments

ASJC Scopus subject areas

  • Software

Cite this

Zhang, Z., Ma, R. T. B., Ding, J., & Yang, Y. (2013). ABACUS: An auction-based approach to cloud service differentiation. In Proceedings of the IEEE International Conference on Cloud Engineering, IC2E 2013 (pp. 292-301). [6529296] https://doi.org/10.1109/IC2E.2013.43

ABACUS : An auction-based approach to cloud service differentiation. / Zhang, Zhenjie; Ma, Richard T B; Ding, Jianbing; Yang, Yin.

Proceedings of the IEEE International Conference on Cloud Engineering, IC2E 2013. 2013. p. 292-301 6529296.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Zhang, Z, Ma, RTB, Ding, J & Yang, Y 2013, ABACUS: An auction-based approach to cloud service differentiation. in Proceedings of the IEEE International Conference on Cloud Engineering, IC2E 2013., 6529296, pp. 292-301, 1st IEEE International Conference on Cloud Engineering, IC2E 2013, San Francisco, CA, United States, 25/3/13. https://doi.org/10.1109/IC2E.2013.43
Zhang Z, Ma RTB, Ding J, Yang Y. ABACUS: An auction-based approach to cloud service differentiation. In Proceedings of the IEEE International Conference on Cloud Engineering, IC2E 2013. 2013. p. 292-301. 6529296 https://doi.org/10.1109/IC2E.2013.43
Zhang, Zhenjie ; Ma, Richard T B ; Ding, Jianbing ; Yang, Yin. / ABACUS : An auction-based approach to cloud service differentiation. Proceedings of the IEEE International Conference on Cloud Engineering, IC2E 2013. 2013. pp. 292-301
@inproceedings{beb1c37b14544c819628277a1446f3eb,
title = "ABACUS: An auction-based approach to cloud service differentiation",
abstract = "The emergence of the cloud computing paradigm has greatly enabled innovative service models, such as Platform as a Service (PaaS), and distributed computing frameworks, such as MapReduce. However, most existing cloud systems fail to distinguish users with different preferences, or jobs of different natures. Consequently, they are unable to provide service differentiation, leading to inefficient allocations of cloud resources. Moreover, contentions on the resources exacerbate this inefficiency, when prioritizing crucial jobs is necessary, but impossible. Motivated by this, we propose Abacus, a generic resource management framework addressing this problem. Abacus interacts with users through an auction mechanism, which allows users to specify their priorities using budgets, and job characteristics via utility functions. Based on this information, Abacus computes the optimal allocation and scheduling of resources. Meanwhile, the auction mechanism in Abacus possesses important properties including incentive compatibility (i.e., the users' best strategy is to simply bid their true budgets and job utilities) and monotonicity (i.e., users are motivated to increase their budgets in order to receive better services). In addition, when the user is unclear about her utility function, Abacus automatically learns this function based on statistics of her previous jobs. An extensive set of experiments, running on Hadoop, demonstrate the high performance and other desirable properties of Abacus.",
author = "Zhenjie Zhang and Ma, {Richard T B} and Jianbing Ding and Yin Yang",
year = "2013",
doi = "10.1109/IC2E.2013.43",
language = "English",
isbn = "9780769549453",
pages = "292--301",
booktitle = "Proceedings of the IEEE International Conference on Cloud Engineering, IC2E 2013",

}

TY - GEN

T1 - ABACUS

T2 - An auction-based approach to cloud service differentiation

AU - Zhang, Zhenjie

AU - Ma, Richard T B

AU - Ding, Jianbing

AU - Yang, Yin

PY - 2013

Y1 - 2013

N2 - The emergence of the cloud computing paradigm has greatly enabled innovative service models, such as Platform as a Service (PaaS), and distributed computing frameworks, such as MapReduce. However, most existing cloud systems fail to distinguish users with different preferences, or jobs of different natures. Consequently, they are unable to provide service differentiation, leading to inefficient allocations of cloud resources. Moreover, contentions on the resources exacerbate this inefficiency, when prioritizing crucial jobs is necessary, but impossible. Motivated by this, we propose Abacus, a generic resource management framework addressing this problem. Abacus interacts with users through an auction mechanism, which allows users to specify their priorities using budgets, and job characteristics via utility functions. Based on this information, Abacus computes the optimal allocation and scheduling of resources. Meanwhile, the auction mechanism in Abacus possesses important properties including incentive compatibility (i.e., the users' best strategy is to simply bid their true budgets and job utilities) and monotonicity (i.e., users are motivated to increase their budgets in order to receive better services). In addition, when the user is unclear about her utility function, Abacus automatically learns this function based on statistics of her previous jobs. An extensive set of experiments, running on Hadoop, demonstrate the high performance and other desirable properties of Abacus.

AB - The emergence of the cloud computing paradigm has greatly enabled innovative service models, such as Platform as a Service (PaaS), and distributed computing frameworks, such as MapReduce. However, most existing cloud systems fail to distinguish users with different preferences, or jobs of different natures. Consequently, they are unable to provide service differentiation, leading to inefficient allocations of cloud resources. Moreover, contentions on the resources exacerbate this inefficiency, when prioritizing crucial jobs is necessary, but impossible. Motivated by this, we propose Abacus, a generic resource management framework addressing this problem. Abacus interacts with users through an auction mechanism, which allows users to specify their priorities using budgets, and job characteristics via utility functions. Based on this information, Abacus computes the optimal allocation and scheduling of resources. Meanwhile, the auction mechanism in Abacus possesses important properties including incentive compatibility (i.e., the users' best strategy is to simply bid their true budgets and job utilities) and monotonicity (i.e., users are motivated to increase their budgets in order to receive better services). In addition, when the user is unclear about her utility function, Abacus automatically learns this function based on statistics of her previous jobs. An extensive set of experiments, running on Hadoop, demonstrate the high performance and other desirable properties of Abacus.

UR - http://www.scopus.com/inward/record.url?scp=84881186979&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84881186979&partnerID=8YFLogxK

U2 - 10.1109/IC2E.2013.43

DO - 10.1109/IC2E.2013.43

M3 - Conference contribution

AN - SCOPUS:84881186979

SN - 9780769549453

SP - 292

EP - 301

BT - Proceedings of the IEEE International Conference on Cloud Engineering, IC2E 2013

ER -