On the accuracy and scalability of intensive I/O workload replay

Alireza Haghdoost; Weiping He; Jerry Fredin; David H.C. Du

On the accuracy and scalability of intensive I/O workload replay

Alireza Haghdoost, Weiping He, Jerry Fredin, David H.C. Du

Computer Science and Engineering

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

16 Scopus citations

Abstract

We introduce a replay tool that can be used to replay captured I/O workloads for performance evaluation of high-performance storage systems. We study several sources in the stock operating system that introduce the uncertainty of replaying a workload. Based on the remedies of these findings, we design and develop a new replay tool called hfplayer that can more accurately replay intensive block I/O workloads in a similar unscaled environment. However, to replay a given workload trace in a scaled environment, the dependency between I/O requests becomes crucial. Therefore, we propose a heuristic way of speculating I/O dependencies in a block I/O trace. Using the generated dependency graph, hfplayer is capable of replaying the I/O workload in a scaled environment. We evaluate hfplayer with a wide range of workloads using several accuracy metrics and find that it produces better accuracy when compared with two exiting available replay tools.

Original language	English (US)
Title of host publication	Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017
Publisher	USENIX Association
Pages	315-327
Number of pages	13
ISBN (Electronic)	9781931971362
State	Published - 2017
Event	15th USENIX Conference on File and Storage Technologies, FAST 2017 - Santa Clara, United States Duration: Feb 27 2017 → Mar 2 2017

Publication series

Name	Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017

Conference

Conference	15th USENIX Conference on File and Storage Technologies, FAST 2017
Country/Territory	United States
City	Santa Clara
Period	2/27/17 → 3/2/17

Bibliographical note

Funding Information:
We thank our shepherd, Remzi Arpaci-Dusseau and Matias Bjørling, and the anonymous reviewers for their comments and suggestions. This work has been supported by NSF I/UCRC Center for Research in Intelligent Storage (CRIS) and the National Science Foundation (NSF) under awards 130523, 1439622, and 1525617 as well as the support from NetApp.

Publisher Copyright:
© Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017. All rights reserved.

Find It

Find It at U of M Libraries

Cite this

On the accuracy and scalability of intensive I/O workload replay. / Haghdoost, Alireza; He, Weiping; Fredin, Jerry et al.

Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017. USENIX Association, 2017. p. 315-327 (Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017).

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

Haghdoost, A, He, W, Fredin, J & Du, DHC 2017, On the accuracy and scalability of intensive I/O workload replay. in Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017. Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017, USENIX Association, pp. 315-327, 15th USENIX Conference on File and Storage Technologies, FAST 2017, Santa Clara, United States, 2/27/17.

@inproceedings{e47a20b0af53435584a967e964226492,

title = "On the accuracy and scalability of intensive I/O workload replay",

abstract = "We introduce a replay tool that can be used to replay captured I/O workloads for performance evaluation of high-performance storage systems. We study several sources in the stock operating system that introduce the uncertainty of replaying a workload. Based on the remedies of these findings, we design and develop a new replay tool called hfplayer that can more accurately replay intensive block I/O workloads in a similar unscaled environment. However, to replay a given workload trace in a scaled environment, the dependency between I/O requests becomes crucial. Therefore, we propose a heuristic way of speculating I/O dependencies in a block I/O trace. Using the generated dependency graph, hfplayer is capable of replaying the I/O workload in a scaled environment. We evaluate hfplayer with a wide range of workloads using several accuracy metrics and find that it produces better accuracy when compared with two exiting available replay tools.",

author = "Alireza Haghdoost and Weiping He and Jerry Fredin and Du, {David H.C.}",

note = "Funding Information: We thank our shepherd, Remzi Arpaci-Dusseau and Matias Bj{\o}rling, and the anonymous reviewers for their comments and suggestions. This work has been supported by NSF I/UCRC Center for Research in Intelligent Storage (CRIS) and the National Science Foundation (NSF) under awards 130523, 1439622, and 1525617 as well as the support from NetApp. Publisher Copyright: {\textcopyright} Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017. All rights reserved.; 15th USENIX Conference on File and Storage Technologies, FAST 2017 ; Conference date: 27-02-2017 Through 02-03-2017",

year = "2017",

language = "English (US)",

series = "Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017",

publisher = "USENIX Association",

pages = "315--327",

booktitle = "Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017",

}

TY - GEN

T1 - On the accuracy and scalability of intensive I/O workload replay

AU - Haghdoost, Alireza

AU - He, Weiping

AU - Fredin, Jerry

AU - Du, David H.C.

N1 - Funding Information: We thank our shepherd, Remzi Arpaci-Dusseau and Matias Bjørling, and the anonymous reviewers for their comments and suggestions. This work has been supported by NSF I/UCRC Center for Research in Intelligent Storage (CRIS) and the National Science Foundation (NSF) under awards 130523, 1439622, and 1525617 as well as the support from NetApp. Publisher Copyright: © Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017. All rights reserved.

PY - 2017

Y1 - 2017

N2 - We introduce a replay tool that can be used to replay captured I/O workloads for performance evaluation of high-performance storage systems. We study several sources in the stock operating system that introduce the uncertainty of replaying a workload. Based on the remedies of these findings, we design and develop a new replay tool called hfplayer that can more accurately replay intensive block I/O workloads in a similar unscaled environment. However, to replay a given workload trace in a scaled environment, the dependency between I/O requests becomes crucial. Therefore, we propose a heuristic way of speculating I/O dependencies in a block I/O trace. Using the generated dependency graph, hfplayer is capable of replaying the I/O workload in a scaled environment. We evaluate hfplayer with a wide range of workloads using several accuracy metrics and find that it produces better accuracy when compared with two exiting available replay tools.

AB - We introduce a replay tool that can be used to replay captured I/O workloads for performance evaluation of high-performance storage systems. We study several sources in the stock operating system that introduce the uncertainty of replaying a workload. Based on the remedies of these findings, we design and develop a new replay tool called hfplayer that can more accurately replay intensive block I/O workloads in a similar unscaled environment. However, to replay a given workload trace in a scaled environment, the dependency between I/O requests becomes crucial. Therefore, we propose a heuristic way of speculating I/O dependencies in a block I/O trace. Using the generated dependency graph, hfplayer is capable of replaying the I/O workload in a scaled environment. We evaluate hfplayer with a wide range of workloads using several accuracy metrics and find that it produces better accuracy when compared with two exiting available replay tools.

UR - http://www.scopus.com/inward/record.url?scp=85031714197&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85031714197&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85031714197

T3 - Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017

SP - 315

EP - 327

BT - Proceedings of the 15th USENIX Conference on File and Storage Technologies, FAST 2017

PB - USENIX Association

T2 - 15th USENIX Conference on File and Storage Technologies, FAST 2017

Y2 - 27 February 2017 through 2 March 2017

ER -

On the accuracy and scalability of intensive I/O workload replay

Abstract

Publication series

Conference

Bibliographical note

Find It

Other files and links

Fingerprint

Cite this