DepCon: Achieving Network SLO for High Performance Clouds

Eunsook Kim; Kyungwoon Lee; Chuck Yoo

doi:10.1007/978-3-031-06156-1_27

DepCon: Achieving Network SLO for High Performance Clouds

Eunsook Kim, Kyungwoon Lee, Chuck Yoo

Department of Computer Science and Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

As containers run in a distributed manner in clouds, it is important to satisfy network service level objectives (SLOs) for containers. In addition, it has been known that containers utilize more CPU resources for network processing than native processes because of the long networking stack of containers. Thus, for achieving network SLOs, containers require sufficient CPU resources as well as network resources, which we call inter-resource dependency. However, existing cloud schedulers have limitations in that they do not take CPU into account for the network SLO. This paper proposes DepCon that controls CPU resources for containers to satisfy network SLOs. DepCon consists of DepCon scheduler that works in the cloud-level and DepCon agent at each node. We implement DepCon in the most popular container orchestration platform, Kubernetes. Our evaluation results show that DepCon reduces the network performance variance by 20 times while improving the network throughput by 40%. In addition, DepCon reduces the scheduling overhead by 20 times compared to the representative multi-resource scheduling technique like DRF.

Original language	English
Title of host publication	Euro-Par 2021
Subtitle of host publication	Parallel Processing Workshops - Euro-Par 2021 International Workshops, 2021, Revised Selected Papers
Editors	Ricardo Chaves, Dora B. Heras, Aleksandar Ilic, Didem Unat, Rosa M. Badia, Andrea Bracciali, Patrick Diehl, Anshu Dubey, Oh Sangyoon, Stephen L. Scott, Laura Ricci
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	339-351
Number of pages	13
ISBN (Print)	9783031061554
DOIs	https://doi.org/10.1007/978-3-031-06156-1_27
Publication status	Published - 2022
Event	27th International Conference on Parallel and Distributed Computing, Euro-Par 2021 - Virtual, Online Duration: 2021 Aug 30 → 2021 Aug 31

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	13098 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	27th International Conference on Parallel and Distributed Computing, Euro-Par 2021
City	Virtual, Online
Period	21/8/30 → 21/8/31

Bibliographical note

Publisher Copyright:
© 2022, Springer Nature Switzerland AG.

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-031-06156-1_27

Cite this

Kim, E., Lee, K., & Yoo, C. (2022). DepCon: Achieving Network SLO for High Performance Clouds. In R. Chaves, D. B. Heras, A. Ilic, D. Unat, R. M. Badia, A. Bracciali, P. Diehl, A. Dubey, O. Sangyoon, S. L. Scott, & L. Ricci (Eds.), Euro-Par 2021: Parallel Processing Workshops - Euro-Par 2021 International Workshops, 2021, Revised Selected Papers (pp. 339-351). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13098 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-06156-1_27

DepCon: Achieving Network SLO for High Performance Clouds. / Kim, Eunsook; Lee, Kyungwoon; Yoo, Chuck.
Euro-Par 2021: Parallel Processing Workshops - Euro-Par 2021 International Workshops, 2021, Revised Selected Papers. ed. / Ricardo Chaves; Dora B. Heras; Aleksandar Ilic; Didem Unat; Rosa M. Badia; Andrea Bracciali; Patrick Diehl; Anshu Dubey; Oh Sangyoon; Stephen L. Scott; Laura Ricci. Springer Science and Business Media Deutschland GmbH, 2022. p. 339-351 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13098 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Kim, E, Lee, K & Yoo, C 2022, DepCon: Achieving Network SLO for High Performance Clouds. in R Chaves, D B. Heras, A Ilic, D Unat, RM Badia, A Bracciali, P Diehl, A Dubey, O Sangyoon, S L. Scott & L Ricci (eds), Euro-Par 2021: Parallel Processing Workshops - Euro-Par 2021 International Workshops, 2021, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 13098 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 339-351, 27th International Conference on Parallel and Distributed Computing, Euro-Par 2021, Virtual, Online, 21/8/30. https://doi.org/10.1007/978-3-031-06156-1_27

Kim E, Lee K, Yoo C. DepCon: Achieving Network SLO for High Performance Clouds. In Chaves R, B. Heras D, Ilic A, Unat D, Badia RM, Bracciali A, Diehl P, Dubey A, Sangyoon O, L. Scott S, Ricci L, editors, Euro-Par 2021: Parallel Processing Workshops - Euro-Par 2021 International Workshops, 2021, Revised Selected Papers. Springer Science and Business Media Deutschland GmbH. 2022. p. 339-351. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-06156-1_27

Kim, Eunsook ; Lee, Kyungwoon ; Yoo, Chuck. / DepCon : Achieving Network SLO for High Performance Clouds. Euro-Par 2021: Parallel Processing Workshops - Euro-Par 2021 International Workshops, 2021, Revised Selected Papers. editor / Ricardo Chaves ; Dora B. Heras ; Aleksandar Ilic ; Didem Unat ; Rosa M. Badia ; Andrea Bracciali ; Patrick Diehl ; Anshu Dubey ; Oh Sangyoon ; Stephen L. Scott ; Laura Ricci. Springer Science and Business Media Deutschland GmbH, 2022. pp. 339-351 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{113f9542a4c1494e8dc44ceb8d5e6082,

title = "DepCon: Achieving Network SLO for High Performance Clouds",

abstract = "As containers run in a distributed manner in clouds, it is important to satisfy network service level objectives (SLOs) for containers. In addition, it has been known that containers utilize more CPU resources for network processing than native processes because of the long networking stack of containers. Thus, for achieving network SLOs, containers require sufficient CPU resources as well as network resources, which we call inter-resource dependency. However, existing cloud schedulers have limitations in that they do not take CPU into account for the network SLO. This paper proposes DepCon that controls CPU resources for containers to satisfy network SLOs. DepCon consists of DepCon scheduler that works in the cloud-level and DepCon agent at each node. We implement DepCon in the most popular container orchestration platform, Kubernetes. Our evaluation results show that DepCon reduces the network performance variance by 20 times while improving the network throughput by 40%. In addition, DepCon reduces the scheduling overhead by 20 times compared to the representative multi-resource scheduling technique like DRF.",

author = "Eunsook Kim and Kyungwoon Lee and Chuck Yoo",

note = "Publisher Copyright: {\textcopyright} 2022, Springer Nature Switzerland AG.; 27th International Conference on Parallel and Distributed Computing, Euro-Par 2021 ; Conference date: 30-08-2021 Through 31-08-2021",

year = "2022",

doi = "10.1007/978-3-031-06156-1_27",

language = "English",

isbn = "9783031061554",

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

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "339--351",

editor = "Ricardo Chaves and {B. Heras}, Dora and Aleksandar Ilic and Didem Unat and Badia, {Rosa M.} and Andrea Bracciali and Patrick Diehl and Anshu Dubey and Oh Sangyoon and {L. Scott}, Stephen and Laura Ricci",

booktitle = "Euro-Par 2021",

address = "Germany",

}

TY - GEN

T1 - DepCon

T2 - 27th International Conference on Parallel and Distributed Computing, Euro-Par 2021

AU - Kim, Eunsook

AU - Lee, Kyungwoon

AU - Yoo, Chuck

PY - 2022

Y1 - 2022

N2 - As containers run in a distributed manner in clouds, it is important to satisfy network service level objectives (SLOs) for containers. In addition, it has been known that containers utilize more CPU resources for network processing than native processes because of the long networking stack of containers. Thus, for achieving network SLOs, containers require sufficient CPU resources as well as network resources, which we call inter-resource dependency. However, existing cloud schedulers have limitations in that they do not take CPU into account for the network SLO. This paper proposes DepCon that controls CPU resources for containers to satisfy network SLOs. DepCon consists of DepCon scheduler that works in the cloud-level and DepCon agent at each node. We implement DepCon in the most popular container orchestration platform, Kubernetes. Our evaluation results show that DepCon reduces the network performance variance by 20 times while improving the network throughput by 40%. In addition, DepCon reduces the scheduling overhead by 20 times compared to the representative multi-resource scheduling technique like DRF.

AB - As containers run in a distributed manner in clouds, it is important to satisfy network service level objectives (SLOs) for containers. In addition, it has been known that containers utilize more CPU resources for network processing than native processes because of the long networking stack of containers. Thus, for achieving network SLOs, containers require sufficient CPU resources as well as network resources, which we call inter-resource dependency. However, existing cloud schedulers have limitations in that they do not take CPU into account for the network SLO. This paper proposes DepCon that controls CPU resources for containers to satisfy network SLOs. DepCon consists of DepCon scheduler that works in the cloud-level and DepCon agent at each node. We implement DepCon in the most popular container orchestration platform, Kubernetes. Our evaluation results show that DepCon reduces the network performance variance by 20 times while improving the network throughput by 40%. In addition, DepCon reduces the scheduling overhead by 20 times compared to the representative multi-resource scheduling technique like DRF.

UR - http://www.scopus.com/inward/record.url?scp=85133020732&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-06156-1_27

DO - 10.1007/978-3-031-06156-1_27

M3 - Conference contribution

AN - SCOPUS:85133020732

SN - 9783031061554

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

SP - 339

EP - 351

BT - Euro-Par 2021

A2 - Chaves, Ricardo

A2 - B. Heras, Dora

A2 - Ilic, Aleksandar

A2 - Unat, Didem

A2 - Badia, Rosa M.

A2 - Bracciali, Andrea

A2 - Diehl, Patrick

A2 - Dubey, Anshu

A2 - Sangyoon, Oh

A2 - L. Scott, Stephen

A2 - Ricci, Laura

PB - Springer Science and Business Media Deutschland GmbH

Y2 - 30 August 2021 through 31 August 2021

ER -

DepCon: Achieving Network SLO for High Performance Clouds

Abstract

Publication series

Conference

Bibliographical note

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this