Autonomic power & performance management for large-scale data centers

Bithika Khargharia, Salim A Hariri, Ferenc Szidarovszky, Manal Houri, Hesham El-Rewini, Samee Ullah Khan, Ishfaq Ahmad, Mazin S. Yousif

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

24 Citations (Scopus)

Abstract

With the rapid growth of servers and applications spurred by the Internet, the power consumption of servers has become critically important and must be efficiently managed. High energy consumption also translates into excessive heat dissipation which in turn, increases cooling costs and causes servers to become more prone to failure. This paper presents a theoretical and experimental framework and general methodology for hierarchical autonomic power & performance management in high performance distributed data centers. We optimize for power & performance (performance/watt) at each level of the hierarchy while maintaining scalability. We adopt mathematicallyrigorous optimization approach to provide the application with the required amount of memory at runtime. This enables us to transition the unused memory capacity to a low power state. Our experimental results show a maximum performance/watt improvement of 88.48% compared to traditional techniques. We also present preliminary results of using Game Theory to optimize performance/watt at the cluster level of a data center. Our cooperative technique reduces the power consumption by 65% when compared to traditional techniques (min-min heuristic).

Original languageEnglish (US)
Title of host publicationProceedings - 21st International Parallel and Distributed Processing Symposium, IPDPS 2007; Abstracts and CD-ROM
DOIs
StatePublished - 2007
Externally publishedYes
Event21st International Parallel and Distributed Processing Symposium, IPDPS 2007 - Long Beach, CA, United States
Duration: Mar 26 2007Mar 30 2007

Other

Other21st International Parallel and Distributed Processing Symposium, IPDPS 2007
CountryUnited States
CityLong Beach, CA
Period3/26/073/30/07

Fingerprint

Performance Management
Data Center
Servers
Server
Electric power utilization
Power Consumption
Data storage equipment
Game theory
Optimise
Heat losses
Scalability
Energy utilization
Game Theory
Internet
Cooling
Energy Consumption
High Energy
Dissipation
High Performance
Heat

ASJC Scopus subject areas

  • Hardware and Architecture
  • Software
  • Mathematics(all)

Cite this

Khargharia, B., Hariri, S. A., Szidarovszky, F., Houri, M., El-Rewini, H., Khan, S. U., ... Yousif, M. S. (2007). Autonomic power & performance management for large-scale data centers. In Proceedings - 21st International Parallel and Distributed Processing Symposium, IPDPS 2007; Abstracts and CD-ROM [4228238] https://doi.org/10.1109/IPDPS.2007.370510

Autonomic power & performance management for large-scale data centers. / Khargharia, Bithika; Hariri, Salim A; Szidarovszky, Ferenc; Houri, Manal; El-Rewini, Hesham; Khan, Samee Ullah; Ahmad, Ishfaq; Yousif, Mazin S.

Proceedings - 21st International Parallel and Distributed Processing Symposium, IPDPS 2007; Abstracts and CD-ROM. 2007. 4228238.

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

Khargharia, B, Hariri, SA, Szidarovszky, F, Houri, M, El-Rewini, H, Khan, SU, Ahmad, I & Yousif, MS 2007, Autonomic power & performance management for large-scale data centers. in Proceedings - 21st International Parallel and Distributed Processing Symposium, IPDPS 2007; Abstracts and CD-ROM., 4228238, 21st International Parallel and Distributed Processing Symposium, IPDPS 2007, Long Beach, CA, United States, 3/26/07. https://doi.org/10.1109/IPDPS.2007.370510
Khargharia B, Hariri SA, Szidarovszky F, Houri M, El-Rewini H, Khan SU et al. Autonomic power & performance management for large-scale data centers. In Proceedings - 21st International Parallel and Distributed Processing Symposium, IPDPS 2007; Abstracts and CD-ROM. 2007. 4228238 https://doi.org/10.1109/IPDPS.2007.370510
Khargharia, Bithika ; Hariri, Salim A ; Szidarovszky, Ferenc ; Houri, Manal ; El-Rewini, Hesham ; Khan, Samee Ullah ; Ahmad, Ishfaq ; Yousif, Mazin S. / Autonomic power & performance management for large-scale data centers. Proceedings - 21st International Parallel and Distributed Processing Symposium, IPDPS 2007; Abstracts and CD-ROM. 2007.
@inproceedings{2d079030b7b944409154c9c015a95f95,
title = "Autonomic power & performance management for large-scale data centers",
abstract = "With the rapid growth of servers and applications spurred by the Internet, the power consumption of servers has become critically important and must be efficiently managed. High energy consumption also translates into excessive heat dissipation which in turn, increases cooling costs and causes servers to become more prone to failure. This paper presents a theoretical and experimental framework and general methodology for hierarchical autonomic power & performance management in high performance distributed data centers. We optimize for power & performance (performance/watt) at each level of the hierarchy while maintaining scalability. We adopt mathematicallyrigorous optimization approach to provide the application with the required amount of memory at runtime. This enables us to transition the unused memory capacity to a low power state. Our experimental results show a maximum performance/watt improvement of 88.48{\%} compared to traditional techniques. We also present preliminary results of using Game Theory to optimize performance/watt at the cluster level of a data center. Our cooperative technique reduces the power consumption by 65{\%} when compared to traditional techniques (min-min heuristic).",
author = "Bithika Khargharia and Hariri, {Salim A} and Ferenc Szidarovszky and Manal Houri and Hesham El-Rewini and Khan, {Samee Ullah} and Ishfaq Ahmad and Yousif, {Mazin S.}",
year = "2007",
doi = "10.1109/IPDPS.2007.370510",
language = "English (US)",
isbn = "1424409101",
booktitle = "Proceedings - 21st International Parallel and Distributed Processing Symposium, IPDPS 2007; Abstracts and CD-ROM",

}

TY - GEN

T1 - Autonomic power & performance management for large-scale data centers

AU - Khargharia, Bithika

AU - Hariri, Salim A

AU - Szidarovszky, Ferenc

AU - Houri, Manal

AU - El-Rewini, Hesham

AU - Khan, Samee Ullah

AU - Ahmad, Ishfaq

AU - Yousif, Mazin S.

PY - 2007

Y1 - 2007

N2 - With the rapid growth of servers and applications spurred by the Internet, the power consumption of servers has become critically important and must be efficiently managed. High energy consumption also translates into excessive heat dissipation which in turn, increases cooling costs and causes servers to become more prone to failure. This paper presents a theoretical and experimental framework and general methodology for hierarchical autonomic power & performance management in high performance distributed data centers. We optimize for power & performance (performance/watt) at each level of the hierarchy while maintaining scalability. We adopt mathematicallyrigorous optimization approach to provide the application with the required amount of memory at runtime. This enables us to transition the unused memory capacity to a low power state. Our experimental results show a maximum performance/watt improvement of 88.48% compared to traditional techniques. We also present preliminary results of using Game Theory to optimize performance/watt at the cluster level of a data center. Our cooperative technique reduces the power consumption by 65% when compared to traditional techniques (min-min heuristic).

AB - With the rapid growth of servers and applications spurred by the Internet, the power consumption of servers has become critically important and must be efficiently managed. High energy consumption also translates into excessive heat dissipation which in turn, increases cooling costs and causes servers to become more prone to failure. This paper presents a theoretical and experimental framework and general methodology for hierarchical autonomic power & performance management in high performance distributed data centers. We optimize for power & performance (performance/watt) at each level of the hierarchy while maintaining scalability. We adopt mathematicallyrigorous optimization approach to provide the application with the required amount of memory at runtime. This enables us to transition the unused memory capacity to a low power state. Our experimental results show a maximum performance/watt improvement of 88.48% compared to traditional techniques. We also present preliminary results of using Game Theory to optimize performance/watt at the cluster level of a data center. Our cooperative technique reduces the power consumption by 65% when compared to traditional techniques (min-min heuristic).

UR - http://www.scopus.com/inward/record.url?scp=34548753308&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34548753308&partnerID=8YFLogxK

U2 - 10.1109/IPDPS.2007.370510

DO - 10.1109/IPDPS.2007.370510

M3 - Conference contribution

SN - 1424409101

SN - 9781424409105

BT - Proceedings - 21st International Parallel and Distributed Processing Symposium, IPDPS 2007; Abstracts and CD-ROM

ER -