Evaluating Load Balancing Policies for Performance and Energy-Efficiency

Freek van den Berg, Björn Frits Postema, Boudewijn R.H.M. Haverkort

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

  • 1 Citations

Abstract

Nowadays, more and more increasingly hard computations are performed in challenging fields like weather forecasting, oil and gas exploration, and cryptanalysis. Many of such computations can be implemented using a computer cluster with a large number of servers. Incoming computation requests are then, via a so-called load balancing policy, distributed over the servers to ensure optimal performance. Additionally, being able to switch-off some servers during low period of workload, gives potential to reduced energy consumption. Therefore, load balancing forms, albeit indirectly, a trade-off between performance and energy consumption. In this paper, we introduce a syntax for load-balancing policies to dynamically select a server for each request based on relevant criteria, including the number of jobs queued in servers, power states of servers, and transition delays between power states of servers. To evaluate many policies, we implement two load balancers in: (i) iDSL, a language and tool-chain for evaluating service-oriented systems, and (ii) a simulation framework in AnyLogic. Both implementations are successfully validated by comparison of the results.
LanguageUndefined
Title of host publicationProceedings of the 14th International Workshop on Quantitative Aspects of Programming Languages and Systems, QAPL 2016
PublisherOpen Publishing Association
Pages98-117
Number of pages20
DOIs
StatePublished - Apr 2016
Event14th International Workshop on Quantitative Aspects of Programming Languages and Systems, QAPL 2016 - Eindhoven, Netherlands
Duration: 2 Apr 20163 Apr 2016
Conference number: 14
http://qapl16.doc.ic.ac.uk/

Publication series

NameElectronic Proceedings in Theoretical Computer Science
PublisherOpen Publishing Association
Volume227
ISSN (Print)2075-2180
ISSN (Electronic)2075-2180

Workshop

Workshop14th International Workshop on Quantitative Aspects of Programming Languages and Systems, QAPL 2016
Abbreviated titleAQPL 2016
CountryNetherlands
CityEindhoven
Period2/04/163/04/16
Internet address

Keywords

  • EWI-27339
  • IR-102548
  • METIS-319466

Cite this

van den Berg, F., Postema, B. F., & Haverkort, B. R. H. M. (2016). Evaluating Load Balancing Policies for Performance and Energy-Efficiency. In Proceedings of the 14th International Workshop on Quantitative Aspects of Programming Languages and Systems, QAPL 2016 (pp. 98-117). (Electronic Proceedings in Theoretical Computer Science; Vol. 227). Open Publishing Association. DOI: 10.4204/EPTCS.227.7
van den Berg, Freek ; Postema, Björn Frits ; Haverkort, Boudewijn R.H.M./ Evaluating Load Balancing Policies for Performance and Energy-Efficiency. Proceedings of the 14th International Workshop on Quantitative Aspects of Programming Languages and Systems, QAPL 2016. Open Publishing Association, 2016. pp. 98-117 (Electronic Proceedings in Theoretical Computer Science).
@inproceedings{fc1f1de093f84dc597367ea5a00a35be,
title = "Evaluating Load Balancing Policies for Performance and Energy-Efficiency",
abstract = "Nowadays, more and more increasingly hard computations are performed in challenging fields like weather forecasting, oil and gas exploration, and cryptanalysis. Many of such computations can be implemented using a computer cluster with a large number of servers. Incoming computation requests are then, via a so-called load balancing policy, distributed over the servers to ensure optimal performance. Additionally, being able to switch-off some servers during low period of workload, gives potential to reduced energy consumption. Therefore, load balancing forms, albeit indirectly, a trade-off between performance and energy consumption. In this paper, we introduce a syntax for load-balancing policies to dynamically select a server for each request based on relevant criteria, including the number of jobs queued in servers, power states of servers, and transition delays between power states of servers. To evaluate many policies, we implement two load balancers in: (i) iDSL, a language and tool-chain for evaluating service-oriented systems, and (ii) a simulation framework in AnyLogic. Both implementations are successfully validated by comparison of the results.",
keywords = "EWI-27339, IR-102548, METIS-319466",
author = "{van den Berg}, Freek and Postema, {Bj{\"o}rn Frits} and Haverkort, {Boudewijn R.H.M.}",
note = "eemcs-eprint-27339",
year = "2016",
month = "4",
doi = "10.4204/EPTCS.227.7",
language = "Undefined",
series = "Electronic Proceedings in Theoretical Computer Science",
publisher = "Open Publishing Association",
pages = "98--117",
booktitle = "Proceedings of the 14th International Workshop on Quantitative Aspects of Programming Languages and Systems, QAPL 2016",

}

van den Berg, F, Postema, BF & Haverkort, BRHM 2016, Evaluating Load Balancing Policies for Performance and Energy-Efficiency. in Proceedings of the 14th International Workshop on Quantitative Aspects of Programming Languages and Systems, QAPL 2016. Electronic Proceedings in Theoretical Computer Science, vol. 227, Open Publishing Association, pp. 98-117, 14th International Workshop on Quantitative Aspects of Programming Languages and Systems, QAPL 2016, Eindhoven, Netherlands, 2/04/16. DOI: 10.4204/EPTCS.227.7

Evaluating Load Balancing Policies for Performance and Energy-Efficiency. / van den Berg, Freek; Postema, Björn Frits; Haverkort, Boudewijn R.H.M.

Proceedings of the 14th International Workshop on Quantitative Aspects of Programming Languages and Systems, QAPL 2016. Open Publishing Association, 2016. p. 98-117 (Electronic Proceedings in Theoretical Computer Science; Vol. 227).

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

TY - GEN

T1 - Evaluating Load Balancing Policies for Performance and Energy-Efficiency

AU - van den Berg,Freek

AU - Postema,Björn Frits

AU - Haverkort,Boudewijn R.H.M.

N1 - eemcs-eprint-27339

PY - 2016/4

Y1 - 2016/4

N2 - Nowadays, more and more increasingly hard computations are performed in challenging fields like weather forecasting, oil and gas exploration, and cryptanalysis. Many of such computations can be implemented using a computer cluster with a large number of servers. Incoming computation requests are then, via a so-called load balancing policy, distributed over the servers to ensure optimal performance. Additionally, being able to switch-off some servers during low period of workload, gives potential to reduced energy consumption. Therefore, load balancing forms, albeit indirectly, a trade-off between performance and energy consumption. In this paper, we introduce a syntax for load-balancing policies to dynamically select a server for each request based on relevant criteria, including the number of jobs queued in servers, power states of servers, and transition delays between power states of servers. To evaluate many policies, we implement two load balancers in: (i) iDSL, a language and tool-chain for evaluating service-oriented systems, and (ii) a simulation framework in AnyLogic. Both implementations are successfully validated by comparison of the results.

AB - Nowadays, more and more increasingly hard computations are performed in challenging fields like weather forecasting, oil and gas exploration, and cryptanalysis. Many of such computations can be implemented using a computer cluster with a large number of servers. Incoming computation requests are then, via a so-called load balancing policy, distributed over the servers to ensure optimal performance. Additionally, being able to switch-off some servers during low period of workload, gives potential to reduced energy consumption. Therefore, load balancing forms, albeit indirectly, a trade-off between performance and energy consumption. In this paper, we introduce a syntax for load-balancing policies to dynamically select a server for each request based on relevant criteria, including the number of jobs queued in servers, power states of servers, and transition delays between power states of servers. To evaluate many policies, we implement two load balancers in: (i) iDSL, a language and tool-chain for evaluating service-oriented systems, and (ii) a simulation framework in AnyLogic. Both implementations are successfully validated by comparison of the results.

KW - EWI-27339

KW - IR-102548

KW - METIS-319466

U2 - 10.4204/EPTCS.227.7

DO - 10.4204/EPTCS.227.7

M3 - Conference contribution

T3 - Electronic Proceedings in Theoretical Computer Science

SP - 98

EP - 117

BT - Proceedings of the 14th International Workshop on Quantitative Aspects of Programming Languages and Systems, QAPL 2016

PB - Open Publishing Association

ER -

van den Berg F, Postema BF, Haverkort BRHM. Evaluating Load Balancing Policies for Performance and Energy-Efficiency. In Proceedings of the 14th International Workshop on Quantitative Aspects of Programming Languages and Systems, QAPL 2016. Open Publishing Association. 2016. p. 98-117. (Electronic Proceedings in Theoretical Computer Science). Available from, DOI: 10.4204/EPTCS.227.7