Dynamic Control of Running Servers

Esa Hyytiä; Douglas Down; Pasi Lassila; Samuli Aalto

doi:10.1007/978-3-319-74947-1_9

Dynamic Control of Running Servers

Esa Hyytiä^*, Douglas Down, Pasi Lassila, Samuli Aalto

^*Corresponding author for this work

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

4 Citations (Scopus)

Abstract

Motivated by a data center setting, we study the problem of joint dispatching and server sleep state control in a system consisting of two queues in parallel. Using the theory of Markov decision processes and a novel lookahead approach, we explicitly determine near-optimal control policies that minimize a combination of QoE costs, energy costs, and wear and tear costs due to switching. Guidelines are provided as to when these combined policies are most effective.

Original language	English
Title of host publication	Measurement, Modelling and Evaluation of Computing Systems - 19th International GI/ITG Conference, MMB 2018, Proceedings
Pages	127-141
Number of pages	15
Volume	10740 LNCS
DOIs	https://doi.org/10.1007/978-3-319-74947-1_9
Publication status	Published - 2018
MoE publication type	A4 Article in a conference publication
Event	International GI/ITG Conference on Measurement, Modelling and Evaluation of Computing Systems - Erlangen, Germany Duration: 26 Feb 2018 → 28 Feb 2018 Conference number: 19

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10740 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International GI/ITG Conference on Measurement, Modelling and Evaluation of Computing Systems
Abbreviated title	MMB
Country	Germany
City	Erlangen
Period	26/02/2018 → 28/02/2018

Keywords

Data centers
Dynamic control
Lookahead techniques
Markov decision processes
Queueing systems

Access to Document

10.1007/978-3-319-74947-1_9

Link to publication in Scopus

Fingerprint Dive into the research topics of 'Dynamic Control of Running Servers'. Together they form a unique fingerprint.

Cite this

Hyytiä, E., Down, D., Lassila, P., & Aalto, S. (2018). Dynamic Control of Running Servers. In Measurement, Modelling and Evaluation of Computing Systems - 19th International GI/ITG Conference, MMB 2018, Proceedings (Vol. 10740 LNCS, pp. 127-141). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10740 LNCS). https://doi.org/10.1007/978-3-319-74947-1_9

Hyytiä, Esa ; Down, Douglas ; Lassila, Pasi ; Aalto, Samuli. / Dynamic Control of Running Servers. Measurement, Modelling and Evaluation of Computing Systems - 19th International GI/ITG Conference, MMB 2018, Proceedings. Vol. 10740 LNCS 2018. pp. 127-141 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{0b4a5399020d4489a811093dc2edcb1a,

title = "Dynamic Control of Running Servers",

abstract = "Motivated by a data center setting, we study the problem of joint dispatching and server sleep state control in a system consisting of two queues in parallel. Using the theory of Markov decision processes and a novel lookahead approach, we explicitly determine near-optimal control policies that minimize a combination of QoE costs, energy costs, and wear and tear costs due to switching. Guidelines are provided as to when these combined policies are most effective.",

keywords = "Data centers, Dynamic control, Lookahead techniques, Markov decision processes, Queueing systems",

author = "Esa Hyyti{\"a} and Douglas Down and Pasi Lassila and Samuli Aalto",

year = "2018",

doi = "10.1007/978-3-319-74947-1_9",

language = "English",

isbn = "9783319749464",

volume = "10740 LNCS",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "127--141",

booktitle = "Measurement, Modelling and Evaluation of Computing Systems - 19th International GI/ITG Conference, MMB 2018, Proceedings",

note = "International GI/ITG Conference on Measurement, Modelling and Evaluation of Computing Systems, MMB ; Conference date: 26-02-2018 Through 28-02-2018",

}

Hyytiä, E, Down, D, Lassila, P & Aalto, S 2018, Dynamic Control of Running Servers. in Measurement, Modelling and Evaluation of Computing Systems - 19th International GI/ITG Conference, MMB 2018, Proceedings. vol. 10740 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10740 LNCS, pp. 127-141, International GI/ITG Conference on Measurement, Modelling and Evaluation of Computing Systems, Erlangen, Germany, 26/02/2018. https://doi.org/10.1007/978-3-319-74947-1_9

Dynamic Control of Running Servers. / Hyytiä, Esa; Down, Douglas; Lassila, Pasi ; Aalto, Samuli.

Measurement, Modelling and Evaluation of Computing Systems - 19th International GI/ITG Conference, MMB 2018, Proceedings. Vol. 10740 LNCS 2018. p. 127-141 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10740 LNCS).

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

TY - GEN

T1 - Dynamic Control of Running Servers

AU - Hyytiä, Esa

AU - Down, Douglas

AU - Lassila, Pasi

AU - Aalto, Samuli

N1 - Conference code: 19

PY - 2018

Y1 - 2018

N2 - Motivated by a data center setting, we study the problem of joint dispatching and server sleep state control in a system consisting of two queues in parallel. Using the theory of Markov decision processes and a novel lookahead approach, we explicitly determine near-optimal control policies that minimize a combination of QoE costs, energy costs, and wear and tear costs due to switching. Guidelines are provided as to when these combined policies are most effective.

AB - Motivated by a data center setting, we study the problem of joint dispatching and server sleep state control in a system consisting of two queues in parallel. Using the theory of Markov decision processes and a novel lookahead approach, we explicitly determine near-optimal control policies that minimize a combination of QoE costs, energy costs, and wear and tear costs due to switching. Guidelines are provided as to when these combined policies are most effective.

KW - Data centers

KW - Dynamic control

KW - Lookahead techniques

KW - Markov decision processes

KW - Queueing systems

UR - http://www.scopus.com/inward/record.url?scp=85042538209&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-74947-1_9

DO - 10.1007/978-3-319-74947-1_9

M3 - Conference contribution

AN - SCOPUS:85042538209

SN - 9783319749464

VL - 10740 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 127

EP - 141

BT - Measurement, Modelling and Evaluation of Computing Systems - 19th International GI/ITG Conference, MMB 2018, Proceedings

T2 - International GI/ITG Conference on Measurement, Modelling and Evaluation of Computing Systems

Y2 - 26 February 2018 through 28 February 2018

ER -

Hyytiä E, Down D, Lassila P , Aalto S. Dynamic Control of Running Servers. In Measurement, Modelling and Evaluation of Computing Systems - 19th International GI/ITG Conference, MMB 2018, Proceedings. Vol. 10740 LNCS. 2018. p. 127-141. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-74947-1_9