Machine learning based performance modeling of flash SSDs

Jaehyung Kim; Jinuk Park; Sanghyun Park

doi:10.1145/3132847.3133120

Machine learning based performance modeling of flash SSDs

Jaehyung Kim, Jinuk Park, Sanghyun Park

College of Computing

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

1 Citation (Scopus)

Abstract

Flash memory based solid state drives(SSDs) have alleviated the I/O bottleneck by exploiting its data parallel design. In an enterprise environment, Flash SSD used in the form of a hybrid storage architecture to achieve the better performance with lower cost. In this architecture, I/O load balancing is one of the important factors. However, the internal parallelism distorts the performance measures of the flash SSDs. Despite the criticality of load balancing on I/O intensive environments, these studies have rarely been addressed. In this paper, we examine the effectiveness of applying classification method using machine learning techniques to the I/O saturation estimation by using Linux kernel I/O statistics instead of the utilization measure that is currently used for HDDs. We conclude that machine learning techniques that we employed (Support Vector Machine and LASSO Generalized Linear Model) performs well compared to the existing utilization measure even we cannot collect the internal information of the flash SSDs.1

Original language	English
Title of host publication	CIKM 2017 - Proceedings of the 2017 ACM Conference on Information and Knowledge Management
Publisher	Association for Computing Machinery
Pages	2135-2138
Number of pages	4
ISBN (Electronic)	9781450349185
DOIs	https://doi.org/10.1145/3132847.3133120
Publication status	Published - 2017 Nov 6
Event	26th ACM International Conference on Information and Knowledge Management, CIKM 2017 - Singapore, Singapore Duration: 2017 Nov 6 → 2017 Nov 10

Publication series

Name	International Conference on Information and Knowledge Management, Proceedings
Volume	Part F131841

Other

Other	26th ACM International Conference on Information and Knowledge Management, CIKM 2017
Country/Territory	Singapore
City	Singapore
Period	17/11/6 → 17/11/10

Bibliographical note

Funding Information:
This research was supported by the MSIT(Ministry of Science and ICT), Korea, under the SW Starlab support program(IITP-2017-0-00477) supervised by the IITP(Institute for Information & communications Technology Promotion).

Publisher Copyright:
© 2017 Association for Computing Machinery.

All Science Journal Classification (ASJC) codes

Decision Sciences(all)
Business, Management and Accounting(all)

Access to Document

10.1145/3132847.3133120

Cite this

Kim, J., Park, J., & Park, S. (2017). Machine learning based performance modeling of flash SSDs. In CIKM 2017 - Proceedings of the 2017 ACM Conference on Information and Knowledge Management (pp. 2135-2138). (International Conference on Information and Knowledge Management, Proceedings; Vol. Part F131841). Association for Computing Machinery. https://doi.org/10.1145/3132847.3133120

@inproceedings{6dbcbd479de246acb138e526e6d5864a,

title = "Machine learning based performance modeling of flash SSDs",

abstract = "Flash memory based solid state drives(SSDs) have alleviated the I/O bottleneck by exploiting its data parallel design. In an enterprise environment, Flash SSD used in the form of a hybrid storage architecture to achieve the better performance with lower cost. In this architecture, I/O load balancing is one of the important factors. However, the internal parallelism distorts the performance measures of the flash SSDs. Despite the criticality of load balancing on I/O intensive environments, these studies have rarely been addressed. In this paper, we examine the effectiveness of applying classification method using machine learning techniques to the I/O saturation estimation by using Linux kernel I/O statistics instead of the utilization measure that is currently used for HDDs. We conclude that machine learning techniques that we employed (Support Vector Machine and LASSO Generalized Linear Model) performs well compared to the existing utilization measure even we cannot collect the internal information of the flash SSDs.1",

author = "Jaehyung Kim and Jinuk Park and Sanghyun Park",

note = "Funding Information: This research was supported by the MSIT(Ministry of Science and ICT), Korea, under the SW Starlab support program(IITP-2017-0-00477) supervised by the IITP(Institute for Information & communications Technology Promotion). Publisher Copyright: {\textcopyright} 2017 Association for Computing Machinery.; 26th ACM International Conference on Information and Knowledge Management, CIKM 2017 ; Conference date: 06-11-2017 Through 10-11-2017",

year = "2017",

month = nov,

day = "6",

doi = "10.1145/3132847.3133120",

language = "English",

series = "International Conference on Information and Knowledge Management, Proceedings",

publisher = "Association for Computing Machinery",

pages = "2135--2138",

booktitle = "CIKM 2017 - Proceedings of the 2017 ACM Conference on Information and Knowledge Management",

}

Kim, J, Park, J & Park, S 2017, Machine learning based performance modeling of flash SSDs. in CIKM 2017 - Proceedings of the 2017 ACM Conference on Information and Knowledge Management. International Conference on Information and Knowledge Management, Proceedings, vol. Part F131841, Association for Computing Machinery, pp. 2135-2138, 26th ACM International Conference on Information and Knowledge Management, CIKM 2017, Singapore, Singapore, 17/11/6. https://doi.org/10.1145/3132847.3133120

Machine learning based performance modeling of flash SSDs. / Kim, Jaehyung; Park, Jinuk; Park, Sanghyun.
CIKM 2017 - Proceedings of the 2017 ACM Conference on Information and Knowledge Management. Association for Computing Machinery, 2017. p. 2135-2138 (International Conference on Information and Knowledge Management, Proceedings; Vol. Part F131841).

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

TY - GEN

T1 - Machine learning based performance modeling of flash SSDs

AU - Kim, Jaehyung

AU - Park, Jinuk

AU - Park, Sanghyun

N1 - Funding Information: This research was supported by the MSIT(Ministry of Science and ICT), Korea, under the SW Starlab support program(IITP-2017-0-00477) supervised by the IITP(Institute for Information & communications Technology Promotion). Publisher Copyright: © 2017 Association for Computing Machinery.

PY - 2017/11/6

Y1 - 2017/11/6

N2 - Flash memory based solid state drives(SSDs) have alleviated the I/O bottleneck by exploiting its data parallel design. In an enterprise environment, Flash SSD used in the form of a hybrid storage architecture to achieve the better performance with lower cost. In this architecture, I/O load balancing is one of the important factors. However, the internal parallelism distorts the performance measures of the flash SSDs. Despite the criticality of load balancing on I/O intensive environments, these studies have rarely been addressed. In this paper, we examine the effectiveness of applying classification method using machine learning techniques to the I/O saturation estimation by using Linux kernel I/O statistics instead of the utilization measure that is currently used for HDDs. We conclude that machine learning techniques that we employed (Support Vector Machine and LASSO Generalized Linear Model) performs well compared to the existing utilization measure even we cannot collect the internal information of the flash SSDs.1

AB - Flash memory based solid state drives(SSDs) have alleviated the I/O bottleneck by exploiting its data parallel design. In an enterprise environment, Flash SSD used in the form of a hybrid storage architecture to achieve the better performance with lower cost. In this architecture, I/O load balancing is one of the important factors. However, the internal parallelism distorts the performance measures of the flash SSDs. Despite the criticality of load balancing on I/O intensive environments, these studies have rarely been addressed. In this paper, we examine the effectiveness of applying classification method using machine learning techniques to the I/O saturation estimation by using Linux kernel I/O statistics instead of the utilization measure that is currently used for HDDs. We conclude that machine learning techniques that we employed (Support Vector Machine and LASSO Generalized Linear Model) performs well compared to the existing utilization measure even we cannot collect the internal information of the flash SSDs.1

UR - http://www.scopus.com/inward/record.url?scp=85037329577&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85037329577&partnerID=8YFLogxK

U2 - 10.1145/3132847.3133120

DO - 10.1145/3132847.3133120

M3 - Conference contribution

AN - SCOPUS:85037329577

T3 - International Conference on Information and Knowledge Management, Proceedings

SP - 2135

EP - 2138

BT - CIKM 2017 - Proceedings of the 2017 ACM Conference on Information and Knowledge Management

PB - Association for Computing Machinery

T2 - 26th ACM International Conference on Information and Knowledge Management, CIKM 2017

Y2 - 6 November 2017 through 10 November 2017

ER -

Machine learning based performance modeling of flash SSDs

Abstract

Publication series

Other

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this