DC-Store: Eliminating Noisy Neighbor Containers using Deterministic I/O Performance and Resource Isolation

Miryeong Kwon; Donghyun Gouk; Changrim Lee; Byounggeun Kim; Jooyoung Hwang; Myoungsoo Jung

DC-Store: Eliminating Noisy Neighbor Containers using Deterministic I/O Performance and Resource Isolation

Miryeong Kwon, Donghyun Gouk, Changrim Lee, Byounggeun Kim, Jooyoung Hwang, Myoungsoo Jung

School of Integrated Technology

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

17 Citations (Scopus)

Abstract

We propose DC-store, a storage framework that offers deterministic I/O performance for a multi-container execution environment. DC-store's hardware-level design implements multiple NVM sets on a shared storage pool, each providing a deterministic SSD access time by removing internal resource conflicts. In parallel, software support of DC-Store is aware of the NVM sets and enlightens Linux kernel to isolate noisy neighbor containers, performing page frame reclaiming, from peers. We prototype both hardware and software counterparts of DC-Store and evaluate them in a real system. The evaluation results demonstrate that containerized data-intensive applications on DC-Store exhibit 31% shorter average execution time, on average, compared to those on a baseline system.

Original language	English
Title of host publication	Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020
Publisher	USENIX Association
Pages	183-191
Number of pages	9
ISBN (Electronic)	9781939133120
Publication status	Published - 2020
Event	18th USENIX Conference on File and Storage Technologies, FAST 2020 - Santa Clara, United States Duration: 2020 Feb 25 → 2020 Feb 27

Publication series

Name	Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020

Conference

Conference	18th USENIX Conference on File and Storage Technologies, FAST 2020
Country/Territory	United States
City	Santa Clara
Period	20/2/25 → 20/2/27

Bibliographical note

Funding Information:
This research is mainly supported by NRF 2016R1C182015312, KAIST Start-up package (G01190015) and Samsung grant (G01190271). The authors thank John Bent of Seagate Technology for shepherding their paper. Myoungsoo Jung is the corresponding author.

Publisher Copyright:
Copyright © Proc. of the 18th USENIX Conference on File and Storage Tech., FAST 2020. All rights reserved.

All Science Journal Classification (ASJC) codes

Hardware and Architecture
Software
Computer Networks and Communications

Cite this

Kwon, M., Gouk, D., Lee, C., Kim, B., Hwang, J., & Jung, M. (2020). DC-Store: Eliminating Noisy Neighbor Containers using Deterministic I/O Performance and Resource Isolation. In Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020 (pp. 183-191). (Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020). USENIX Association.

Kwon, Miryeong ; Gouk, Donghyun ; Lee, Changrim et al. / DC-Store : Eliminating Noisy Neighbor Containers using Deterministic I/O Performance and Resource Isolation. Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020. USENIX Association, 2020. pp. 183-191 (Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020).

@inproceedings{4aa9ab848c7346499c0d4597c16646fa,

title = "DC-Store: Eliminating Noisy Neighbor Containers using Deterministic I/O Performance and Resource Isolation",

abstract = "We propose DC-store, a storage framework that offers deterministic I/O performance for a multi-container execution environment. DC-store's hardware-level design implements multiple NVM sets on a shared storage pool, each providing a deterministic SSD access time by removing internal resource conflicts. In parallel, software support of DC-Store is aware of the NVM sets and enlightens Linux kernel to isolate noisy neighbor containers, performing page frame reclaiming, from peers. We prototype both hardware and software counterparts of DC-Store and evaluate them in a real system. The evaluation results demonstrate that containerized data-intensive applications on DC-Store exhibit 31% shorter average execution time, on average, compared to those on a baseline system.",

author = "Miryeong Kwon and Donghyun Gouk and Changrim Lee and Byounggeun Kim and Jooyoung Hwang and Myoungsoo Jung",

note = "Funding Information: This research is mainly supported by NRF 2016R1C182015312, KAIST Start-up package (G01190015) and Samsung grant (G01190271). The authors thank John Bent of Seagate Technology for shepherding their paper. Myoungsoo Jung is the corresponding author. Publisher Copyright: Copyright {\textcopyright} Proc. of the 18th USENIX Conference on File and Storage Tech., FAST 2020. All rights reserved.; 18th USENIX Conference on File and Storage Technologies, FAST 2020 ; Conference date: 25-02-2020 Through 27-02-2020",

year = "2020",

language = "English",

series = "Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020",

publisher = "USENIX Association",

pages = "183--191",

booktitle = "Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020",

}

Kwon, M, Gouk, D, Lee, C, Kim, B, Hwang, J & Jung, M 2020, DC-Store: Eliminating Noisy Neighbor Containers using Deterministic I/O Performance and Resource Isolation. in Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020. Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020, USENIX Association, pp. 183-191, 18th USENIX Conference on File and Storage Technologies, FAST 2020, Santa Clara, United States, 20/2/25.

DC-Store : Eliminating Noisy Neighbor Containers using Deterministic I/O Performance and Resource Isolation. / Kwon, Miryeong; Gouk, Donghyun; Lee, Changrim et al.

Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020. USENIX Association, 2020. p. 183-191 (Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020).

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

TY - GEN

T1 - DC-Store

T2 - 18th USENIX Conference on File and Storage Technologies, FAST 2020

AU - Kwon, Miryeong

AU - Gouk, Donghyun

AU - Lee, Changrim

AU - Kim, Byounggeun

AU - Hwang, Jooyoung

AU - Jung, Myoungsoo

N1 - Funding Information: This research is mainly supported by NRF 2016R1C182015312, KAIST Start-up package (G01190015) and Samsung grant (G01190271). The authors thank John Bent of Seagate Technology for shepherding their paper. Myoungsoo Jung is the corresponding author. Publisher Copyright: Copyright © Proc. of the 18th USENIX Conference on File and Storage Tech., FAST 2020. All rights reserved.

PY - 2020

Y1 - 2020

N2 - We propose DC-store, a storage framework that offers deterministic I/O performance for a multi-container execution environment. DC-store's hardware-level design implements multiple NVM sets on a shared storage pool, each providing a deterministic SSD access time by removing internal resource conflicts. In parallel, software support of DC-Store is aware of the NVM sets and enlightens Linux kernel to isolate noisy neighbor containers, performing page frame reclaiming, from peers. We prototype both hardware and software counterparts of DC-Store and evaluate them in a real system. The evaluation results demonstrate that containerized data-intensive applications on DC-Store exhibit 31% shorter average execution time, on average, compared to those on a baseline system.

AB - We propose DC-store, a storage framework that offers deterministic I/O performance for a multi-container execution environment. DC-store's hardware-level design implements multiple NVM sets on a shared storage pool, each providing a deterministic SSD access time by removing internal resource conflicts. In parallel, software support of DC-Store is aware of the NVM sets and enlightens Linux kernel to isolate noisy neighbor containers, performing page frame reclaiming, from peers. We prototype both hardware and software counterparts of DC-Store and evaluate them in a real system. The evaluation results demonstrate that containerized data-intensive applications on DC-Store exhibit 31% shorter average execution time, on average, compared to those on a baseline system.

UR - http://www.scopus.com/inward/record.url?scp=85091889094&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85091889094&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85091889094

T3 - Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020

SP - 183

EP - 191

BT - Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020

PB - USENIX Association

Y2 - 25 February 2020 through 27 February 2020

ER -

Kwon M, Gouk D, Lee C, Kim B, Hwang J, Jung M. DC-Store: Eliminating Noisy Neighbor Containers using Deterministic I/O Performance and Resource Isolation. In Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020. USENIX Association. 2020. p. 183-191. (Proceedings of the 18th USENIX Conference on File and Storage Technologies, FAST 2020).

DC-Store: Eliminating Noisy Neighbor Containers using Deterministic I/O Performance and Resource Isolation

Abstract

Publication series

Conference

Bibliographical note

All Science Journal Classification (ASJC) codes

Other files and links

Fingerprint

Cite this