Energy-aware server with SRPT scheduling: Analysis and optimization

Misikir Eyob Gebrehiwot; Samuli Aalto; Pasi Lassila

doi:10.1007/978-3-319-43425-4_7

Energy-aware server with SRPT scheduling: Analysis and optimization

Misikir Eyob Gebrehiwot, Samuli Aalto, Pasi Lassila

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

5 Citations (Scopus)

Abstract

We consider the optimal energy-aware control of a single server in a server farm. The server is modeled as an M/G/1 queue with a particular control policy that allows to put the server to a sleep mode to save energy with an additional delay cost, the setup delay, after the server is turned on again. Our main result is the derivation of mean response time for such a system under SRPT scheduling. In particular, we show that the mean response time can be decomposed into two parts: the mean response time of an ordinary M/G/1-SRPT, and an additional penalty term for switching the server to a sleep state. Furthermore, we study the energy-performance optimization of the system and prove that, for the Energy Response time Weighted Sum (ERWS) and Energy Response time Product (ERP) cost metrics, the optimal control either puts the server into a sleep state immediately when it becomes idle or keeps it idling until the next job arrives.

Original language	English
Title of host publication	Quantitative Evaluation of Systems
Subtitle of host publication	13th International Conference, QEST 2016, Quebec City, QC, Canada, August 23-25, 2016, Proceedings
Editors	Gul Agha, Benny Van Houdt
Pages	107-122
Number of pages	16
ISBN (Electronic)	978-3-319-43425-4
DOIs	https://doi.org/10.1007/978-3-319-43425-4_7
Publication status	Published - 3 Aug 2016
MoE publication type	A4 Article in a conference publication
Event	International Conference on Quantitative Evaluation of Systems - Quebec City, Canada Duration: 23 Aug 2016 → 25 Aug 2016 Conference number: 13

Publication series

Name	Lecture notes in computer science
Publisher	Springer
Number	9826
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Conference on Quantitative Evaluation of Systems
Abbreviated title	QEST
Country	Canada
City	Quebec City
Period	23/08/2016 → 25/08/2016

Keywords

Performance-energy trade-off
M/G/1-SRPT
Setup delay

Access to Document

10.1007/978-3-319-43425-4_7

Fingerprint Dive into the research topics of 'Energy-aware server with SRPT scheduling: Analysis and optimization'. Together they form a unique fingerprint.

Projects

1 Finished

Towards Optimal Performance-Energy Trade-off in Server Farms

Aalto, S., Gebrehiwot, M., Hyytiä, E., Lassila, P. & Osti, P.

01/09/2013 → 31/08/2017

Project: Academy of Finland: Other research funding

Cite this

Gebrehiwot, M. E., Aalto, S., & Lassila, P. (2016). Energy-aware server with SRPT scheduling: Analysis and optimization. In G. Agha, & B. Van Houdt (Eds.), Quantitative Evaluation of Systems: 13th International Conference, QEST 2016, Quebec City, QC, Canada, August 23-25, 2016, Proceedings (pp. 107-122). (Lecture notes in computer science; No. 9826). https://doi.org/10.1007/978-3-319-43425-4_7

@inproceedings{8c21588797c44b20963b518538ae13fa,

title = "Energy-aware server with SRPT scheduling: Analysis and optimization",

abstract = "We consider the optimal energy-aware control of a single server in a server farm. The server is modeled as an M/G/1 queue with a particular control policy that allows to put the server to a sleep mode to save energy with an additional delay cost, the setup delay, after the server is turned on again. Our main result is the derivation of mean response time for such a system under SRPT scheduling. In particular, we show that the mean response time can be decomposed into two parts: the mean response time of an ordinary M/G/1-SRPT, and an additional penalty term for switching the server to a sleep state. Furthermore, we study the energy-performance optimization of the system and prove that, for the Energy Response time Weighted Sum (ERWS) and Energy Response time Product (ERP) cost metrics, the optimal control either puts the server into a sleep state immediately when it becomes idle or keeps it idling until the next job arrives.",

keywords = "Performance-energy trade-off, M/G/1-SRPT, Setup delay",

author = "Gebrehiwot, {Misikir Eyob} and Samuli Aalto and Pasi Lassila",

year = "2016",

month = aug,

day = "3",

doi = "10.1007/978-3-319-43425-4_7",

language = "English",

isbn = "978-3-319-43424-7",

series = "Lecture notes in computer science",

publisher = "Springer",

number = "9826",

pages = "107--122",

editor = "Gul Agha and {Van Houdt}, Benny",

booktitle = "Quantitative Evaluation of Systems",

note = "International Conference on Quantitative Evaluation of Systems, QEST ; Conference date: 23-08-2016 Through 25-08-2016",

}

Gebrehiwot, ME, Aalto, S & Lassila, P 2016, Energy-aware server with SRPT scheduling: Analysis and optimization. in G Agha & B Van Houdt (eds), Quantitative Evaluation of Systems: 13th International Conference, QEST 2016, Quebec City, QC, Canada, August 23-25, 2016, Proceedings. Lecture notes in computer science, no. 9826, pp. 107-122, International Conference on Quantitative Evaluation of Systems, Quebec City, Canada, 23/08/2016. https://doi.org/10.1007/978-3-319-43425-4_7

Energy-aware server with SRPT scheduling : Analysis and optimization. / Gebrehiwot, Misikir Eyob; Aalto, Samuli ; Lassila, Pasi.

Quantitative Evaluation of Systems: 13th International Conference, QEST 2016, Quebec City, QC, Canada, August 23-25, 2016, Proceedings. ed. / Gul Agha; Benny Van Houdt. 2016. p. 107-122 (Lecture notes in computer science; No. 9826).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Energy-aware server with SRPT scheduling

T2 - International Conference on Quantitative Evaluation of Systems

AU - Gebrehiwot, Misikir Eyob

AU - Aalto, Samuli

AU - Lassila, Pasi

N1 - Conference code: 13

PY - 2016/8/3

Y1 - 2016/8/3

N2 - We consider the optimal energy-aware control of a single server in a server farm. The server is modeled as an M/G/1 queue with a particular control policy that allows to put the server to a sleep mode to save energy with an additional delay cost, the setup delay, after the server is turned on again. Our main result is the derivation of mean response time for such a system under SRPT scheduling. In particular, we show that the mean response time can be decomposed into two parts: the mean response time of an ordinary M/G/1-SRPT, and an additional penalty term for switching the server to a sleep state. Furthermore, we study the energy-performance optimization of the system and prove that, for the Energy Response time Weighted Sum (ERWS) and Energy Response time Product (ERP) cost metrics, the optimal control either puts the server into a sleep state immediately when it becomes idle or keeps it idling until the next job arrives.

AB - We consider the optimal energy-aware control of a single server in a server farm. The server is modeled as an M/G/1 queue with a particular control policy that allows to put the server to a sleep mode to save energy with an additional delay cost, the setup delay, after the server is turned on again. Our main result is the derivation of mean response time for such a system under SRPT scheduling. In particular, we show that the mean response time can be decomposed into two parts: the mean response time of an ordinary M/G/1-SRPT, and an additional penalty term for switching the server to a sleep state. Furthermore, we study the energy-performance optimization of the system and prove that, for the Energy Response time Weighted Sum (ERWS) and Energy Response time Product (ERP) cost metrics, the optimal control either puts the server into a sleep state immediately when it becomes idle or keeps it idling until the next job arrives.

KW - Performance-energy trade-off

KW - M/G/1-SRPT

KW - Setup delay

U2 - 10.1007/978-3-319-43425-4_7

DO - 10.1007/978-3-319-43425-4_7

M3 - Conference contribution

SN - 978-3-319-43424-7

T3 - Lecture notes in computer science

SP - 107

EP - 122

BT - Quantitative Evaluation of Systems

A2 - Agha, Gul

A2 - Van Houdt, Benny

Y2 - 23 August 2016 through 25 August 2016

ER -

Gebrehiwot ME, Aalto S , Lassila P. Energy-aware server with SRPT scheduling: Analysis and optimization. In Agha G, Van Houdt B, editors, Quantitative Evaluation of Systems: 13th International Conference, QEST 2016, Quebec City, QC, Canada, August 23-25, 2016, Proceedings. 2016. p. 107-122. (Lecture notes in computer science; 9826). https://doi.org/10.1007/978-3-319-43425-4_7