Governor

Autonomic throttling for aggressive idle resource scavenging

Jonathan W. Strickland, Vincent W. Freeh, Xiaosong Ma, Sudharshan S. Vazhkudai

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

12 Citations (Scopus)

Abstract

Scavenging (or resource borrowing) is a common approach used to harness unused resources to perform useful calculations. Since these are volunteer contributions from resource owners, it is vital to reduce the impact of scavenging activities on their native workload to a minimum. To this end, existing impact control systems are either overly conservative in stopping scavenging altogether or inflexible and lack user autonomy to regulate resource usage as in some priority-based techniques. In this paper, we propose a systematic impact control framework for resource scavenging, by quantifying the performance impact a scavenging application incurs on a set of tasks stressing different system resources. For a user-configurable impact threshold, the framework monitors the native workload, determines the dominating native task, and autonomically and adoptively throttles the scavenging application, to bring the impact below target levels. This novel approach has unique benefits of 1) making impact control explicit to resource owners and an easy-to-tune "knob," and 2) adapting to different scavenging applications and native workloads. Our experiments with two scavenging applications, which use resources in very different ways, demonstrate that this framework allows both more aggressive resource scavenging and less impact on native workloads at the same time, compared to a priority-based method. Finally, the framework itself is a lightweight user-level process whose monitoring overhead on native workloads averages as low as 1%.

Original languageEnglish
Title of host publicationProceedings - Second International Conference on Autonomic Computing, ICAC 2005
Pages64-75
Number of pages12
Volume2005
DOIs
Publication statusPublished - 1 Dec 2005
Externally publishedYes
Event2nd International Conference on Autonomic Computing, ICAC 2005 - Seattle, WA, United States
Duration: 13 Jun 200516 Jun 2005

Other

Other2nd International Conference on Autonomic Computing, ICAC 2005
CountryUnited States
CitySeattle, WA
Period13/6/0516/6/05

Fingerprint

Governors
Scavenging
Knobs
Process monitoring
Control systems

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Strickland, J. W., Freeh, V. W., Ma, X., & Vazhkudai, S. S. (2005). Governor: Autonomic throttling for aggressive idle resource scavenging. In Proceedings - Second International Conference on Autonomic Computing, ICAC 2005 (Vol. 2005, pp. 64-75). [1498053] https://doi.org/10.1109/ICAC.2005.31

Governor : Autonomic throttling for aggressive idle resource scavenging. / Strickland, Jonathan W.; Freeh, Vincent W.; Ma, Xiaosong; Vazhkudai, Sudharshan S.

Proceedings - Second International Conference on Autonomic Computing, ICAC 2005. Vol. 2005 2005. p. 64-75 1498053.

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

Strickland, JW, Freeh, VW, Ma, X & Vazhkudai, SS 2005, Governor: Autonomic throttling for aggressive idle resource scavenging. in Proceedings - Second International Conference on Autonomic Computing, ICAC 2005. vol. 2005, 1498053, pp. 64-75, 2nd International Conference on Autonomic Computing, ICAC 2005, Seattle, WA, United States, 13/6/05. https://doi.org/10.1109/ICAC.2005.31
Strickland JW, Freeh VW, Ma X, Vazhkudai SS. Governor: Autonomic throttling for aggressive idle resource scavenging. In Proceedings - Second International Conference on Autonomic Computing, ICAC 2005. Vol. 2005. 2005. p. 64-75. 1498053 https://doi.org/10.1109/ICAC.2005.31
Strickland, Jonathan W. ; Freeh, Vincent W. ; Ma, Xiaosong ; Vazhkudai, Sudharshan S. / Governor : Autonomic throttling for aggressive idle resource scavenging. Proceedings - Second International Conference on Autonomic Computing, ICAC 2005. Vol. 2005 2005. pp. 64-75
@inproceedings{8b84d8e413254e08a9a098cd274a264f,
title = "Governor: Autonomic throttling for aggressive idle resource scavenging",
abstract = "Scavenging (or resource borrowing) is a common approach used to harness unused resources to perform useful calculations. Since these are volunteer contributions from resource owners, it is vital to reduce the impact of scavenging activities on their native workload to a minimum. To this end, existing impact control systems are either overly conservative in stopping scavenging altogether or inflexible and lack user autonomy to regulate resource usage as in some priority-based techniques. In this paper, we propose a systematic impact control framework for resource scavenging, by quantifying the performance impact a scavenging application incurs on a set of tasks stressing different system resources. For a user-configurable impact threshold, the framework monitors the native workload, determines the dominating native task, and autonomically and adoptively throttles the scavenging application, to bring the impact below target levels. This novel approach has unique benefits of 1) making impact control explicit to resource owners and an easy-to-tune {"}knob,{"} and 2) adapting to different scavenging applications and native workloads. Our experiments with two scavenging applications, which use resources in very different ways, demonstrate that this framework allows both more aggressive resource scavenging and less impact on native workloads at the same time, compared to a priority-based method. Finally, the framework itself is a lightweight user-level process whose monitoring overhead on native workloads averages as low as 1{\%}.",
author = "Strickland, {Jonathan W.} and Freeh, {Vincent W.} and Xiaosong Ma and Vazhkudai, {Sudharshan S.}",
year = "2005",
month = "12",
day = "1",
doi = "10.1109/ICAC.2005.31",
language = "English",
isbn = "0769522769",
volume = "2005",
pages = "64--75",
booktitle = "Proceedings - Second International Conference on Autonomic Computing, ICAC 2005",

}

TY - GEN

T1 - Governor

T2 - Autonomic throttling for aggressive idle resource scavenging

AU - Strickland, Jonathan W.

AU - Freeh, Vincent W.

AU - Ma, Xiaosong

AU - Vazhkudai, Sudharshan S.

PY - 2005/12/1

Y1 - 2005/12/1

N2 - Scavenging (or resource borrowing) is a common approach used to harness unused resources to perform useful calculations. Since these are volunteer contributions from resource owners, it is vital to reduce the impact of scavenging activities on their native workload to a minimum. To this end, existing impact control systems are either overly conservative in stopping scavenging altogether or inflexible and lack user autonomy to regulate resource usage as in some priority-based techniques. In this paper, we propose a systematic impact control framework for resource scavenging, by quantifying the performance impact a scavenging application incurs on a set of tasks stressing different system resources. For a user-configurable impact threshold, the framework monitors the native workload, determines the dominating native task, and autonomically and adoptively throttles the scavenging application, to bring the impact below target levels. This novel approach has unique benefits of 1) making impact control explicit to resource owners and an easy-to-tune "knob," and 2) adapting to different scavenging applications and native workloads. Our experiments with two scavenging applications, which use resources in very different ways, demonstrate that this framework allows both more aggressive resource scavenging and less impact on native workloads at the same time, compared to a priority-based method. Finally, the framework itself is a lightweight user-level process whose monitoring overhead on native workloads averages as low as 1%.

AB - Scavenging (or resource borrowing) is a common approach used to harness unused resources to perform useful calculations. Since these are volunteer contributions from resource owners, it is vital to reduce the impact of scavenging activities on their native workload to a minimum. To this end, existing impact control systems are either overly conservative in stopping scavenging altogether or inflexible and lack user autonomy to regulate resource usage as in some priority-based techniques. In this paper, we propose a systematic impact control framework for resource scavenging, by quantifying the performance impact a scavenging application incurs on a set of tasks stressing different system resources. For a user-configurable impact threshold, the framework monitors the native workload, determines the dominating native task, and autonomically and adoptively throttles the scavenging application, to bring the impact below target levels. This novel approach has unique benefits of 1) making impact control explicit to resource owners and an easy-to-tune "knob," and 2) adapting to different scavenging applications and native workloads. Our experiments with two scavenging applications, which use resources in very different ways, demonstrate that this framework allows both more aggressive resource scavenging and less impact on native workloads at the same time, compared to a priority-based method. Finally, the framework itself is a lightweight user-level process whose monitoring overhead on native workloads averages as low as 1%.

UR - http://www.scopus.com/inward/record.url?scp=33745508736&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33745508736&partnerID=8YFLogxK

U2 - 10.1109/ICAC.2005.31

DO - 10.1109/ICAC.2005.31

M3 - Conference contribution

SN - 0769522769

SN - 9780769522760

VL - 2005

SP - 64

EP - 75

BT - Proceedings - Second International Conference on Autonomic Computing, ICAC 2005

ER -