hfplayer: Scalable Replay for Intensive Block I/O Workloads

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

doi:10.1145/3149392

hfplayer: Scalable Replay for Intensive Block I/O Workloads

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

Computer Science and Engineering

Research output: Contribution to journal › Article › peer-review

2 Scopus citations

Abstract

We introduce new methods to replay intensive block I/O workloads more accurately. These methods can be used to reproduce realistic workloads for benchmarking, performance validation, and tuning of a high-performance block storage device/system. In this article, we study several sources in the stock operating system that introduce uncertainty in the workload replay. Based on the remedies of these findings, we design and develop a new replay tool called hfplayer that replays intensive block I/O workloads in a similar unscaled environment with more accuracy. To replay a given workload trace in a scaled environment with faster storage or host server, the dependency between I/O requests becomes crucial since the timing and ordering of I/O requests is expected to change according to these dependencies. Therefore, we propose a heuristic way of speculating I/O dependencies in a block I/O trace. Using the generated dependency graph, hfplayer tries to propagate I/O related performance gains appropriately along the I/O dependency chains and mimics the original application behavior when it executes in a scaled environment with slower or faster storage system and servers. We evaluate hfplayer with a wide range of workloads using several accuracy metrics and find that it produces better accuracy when compared to other replay approaches.

Original language	English (US)
Article number	Y
Journal	ACM Transactions on Storage
Volume	13
Issue number	4
DOIs	https://doi.org/10.1145/3149392
State	Published - Dec 2017

Bibliographical note

Funding Information:
This work was 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 by NetApp. Authors’ addresses: A. Haghdoost, W. He, and D. H. C. Du, 200 Union Street SE Minneapolis, MN 55455; emails: {alireza, weihe, du}@cs.umn.edu; J. Fredin, 3718 N Rock Rd, Wichita, KS 67226; email: [email protected]. Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected]. © 2017 ACM 1553-3077/2017/12-ART39 $15.00 https://doi.org/10.1145/3149392

Publisher Copyright:
© 2017 ACM.

Keywords

I/O workload
Storage area network
Storage performance
Workload replay

Publisher link

10.1145/3149392

Find It

Find It at U of M Libraries

Cite this

@article{d92e9e41903144b0a21558cdf5b182d5,

title = "hfplayer: Scalable Replay for Intensive Block I/O Workloads",

abstract = "We introduce new methods to replay intensive block I/O workloads more accurately. These methods can be used to reproduce realistic workloads for benchmarking, performance validation, and tuning of a high-performance block storage device/system. In this article, we study several sources in the stock operating system that introduce uncertainty in the workload replay. Based on the remedies of these findings, we design and develop a new replay tool called hfplayer that replays intensive block I/O workloads in a similar unscaled environment with more accuracy. To replay a given workload trace in a scaled environment with faster storage or host server, the dependency between I/O requests becomes crucial since the timing and ordering of I/O requests is expected to change according to these dependencies. Therefore, we propose a heuristic way of speculating I/O dependencies in a block I/O trace. Using the generated dependency graph, hfplayer tries to propagate I/O related performance gains appropriately along the I/O dependency chains and mimics the original application behavior when it executes in a scaled environment with slower or faster storage system and servers. We evaluate hfplayer with a wide range of workloads using several accuracy metrics and find that it produces better accuracy when compared to other replay approaches.",

keywords = "I/O workload, Storage area network, Storage performance, Workload replay",

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

note = "Funding Information: This work was 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 by NetApp. Authors{\textquoteright} addresses: A. Haghdoost, W. He, and D. H. C. Du, 200 Union Street SE Minneapolis, MN 55455; emails: {alireza, weihe, du}@cs.umn.edu; J. Fredin, 3718 N Rock Rd, Wichita, KS 67226; email: [email protected]. Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected]. {\textcopyright} 2017 ACM 1553-3077/2017/12-ART39 $15.00 https://doi.org/10.1145/3149392 Publisher Copyright: {\textcopyright} 2017 ACM.",

year = "2017",

month = dec,

doi = "10.1145/3149392",

language = "English (US)",

volume = "13",

journal = "ACM Transactions on Storage",

issn = "1553-3077",

publisher = "Association for Computing Machinery (ACM)",

number = "4",

}

TY - JOUR

T1 - hfplayer

T2 - Scalable Replay for Intensive Block I/O Workloads

AU - Haghdoost, Alireza

AU - He, Weiping

AU - Fredin, Jerry

AU - Du, David H

N1 - Funding Information: This work was 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 by NetApp. Authors’ addresses: A. Haghdoost, W. He, and D. H. C. Du, 200 Union Street SE Minneapolis, MN 55455; emails: {alireza, weihe, du}@cs.umn.edu; J. Fredin, 3718 N Rock Rd, Wichita, KS 67226; email: [email protected]. Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected]. © 2017 ACM 1553-3077/2017/12-ART39 $15.00 https://doi.org/10.1145/3149392 Publisher Copyright: © 2017 ACM.

PY - 2017/12

Y1 - 2017/12

N2 - We introduce new methods to replay intensive block I/O workloads more accurately. These methods can be used to reproduce realistic workloads for benchmarking, performance validation, and tuning of a high-performance block storage device/system. In this article, we study several sources in the stock operating system that introduce uncertainty in the workload replay. Based on the remedies of these findings, we design and develop a new replay tool called hfplayer that replays intensive block I/O workloads in a similar unscaled environment with more accuracy. To replay a given workload trace in a scaled environment with faster storage or host server, the dependency between I/O requests becomes crucial since the timing and ordering of I/O requests is expected to change according to these dependencies. Therefore, we propose a heuristic way of speculating I/O dependencies in a block I/O trace. Using the generated dependency graph, hfplayer tries to propagate I/O related performance gains appropriately along the I/O dependency chains and mimics the original application behavior when it executes in a scaled environment with slower or faster storage system and servers. We evaluate hfplayer with a wide range of workloads using several accuracy metrics and find that it produces better accuracy when compared to other replay approaches.

AB - We introduce new methods to replay intensive block I/O workloads more accurately. These methods can be used to reproduce realistic workloads for benchmarking, performance validation, and tuning of a high-performance block storage device/system. In this article, we study several sources in the stock operating system that introduce uncertainty in the workload replay. Based on the remedies of these findings, we design and develop a new replay tool called hfplayer that replays intensive block I/O workloads in a similar unscaled environment with more accuracy. To replay a given workload trace in a scaled environment with faster storage or host server, the dependency between I/O requests becomes crucial since the timing and ordering of I/O requests is expected to change according to these dependencies. Therefore, we propose a heuristic way of speculating I/O dependencies in a block I/O trace. Using the generated dependency graph, hfplayer tries to propagate I/O related performance gains appropriately along the I/O dependency chains and mimics the original application behavior when it executes in a scaled environment with slower or faster storage system and servers. We evaluate hfplayer with a wide range of workloads using several accuracy metrics and find that it produces better accuracy when compared to other replay approaches.

KW - I/O workload

KW - Storage area network

KW - Storage performance

KW - Workload replay

UR - http://www.scopus.com/inward/record.url?scp=85040307458&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85040307458&partnerID=8YFLogxK

U2 - 10.1145/3149392

DO - 10.1145/3149392

M3 - Article

AN - SCOPUS:85040307458

SN - 1553-3077

VL - 13

JO - ACM Transactions on Storage

JF - ACM Transactions on Storage

IS - 4

M1 - Y

ER -

hfplayer: Scalable Replay for Intensive Block I/O Workloads

Abstract

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this