TrapFetch: A breakpoint-based prefetcher for both launch and run-time

Jiwoong Won; Oseok Kwon; Junhee Ryu; Junbeom Hur; Insup Lee; Kyungtae Kang

doi:10.1109/SMC.2017.8123045

TrapFetch: A breakpoint-based prefetcher for both launch and run-time

Jiwoong Won, Oseok Kwon, Junhee Ryu, Junbeom Hur, Insup Lee, Kyungtae Kang

Department of Computer Science and Engineering

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

3 Citations (Scopus)

Abstract

TrapFetch is trained by monitoring the read requests issued by an application. It detects bursts of disk reads, determines the appropriate addresses at which breakpoints should be inserted in the application and library codes prior to the bursts of reads, and then logs this information with the data requested during the interval between each consecutive pair of breakpoints. When the application and library codes are loaded from the disk into memory, TrapFetch inserts breakpoints at the designated addresses based on the logs. During subsequent runs, TrapFetch is invoked at each breakpoint when it prefetches the corresponding data into the page cache. This approach is effective during both launch and run-time. TrapFetch operates at the user level, thus avoiding interference with other applications. In experiments on five popular applications (FlightGear, SpeedDreams 2, Pillars of Eternity, Eclipse, and VegaStrike), TrapFetch reduced the time for launch by up to 39.7% and time for run-time dataloading by up to 63.7%.

Original language	English
Title of host publication	2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	2766-2771
Number of pages	6
ISBN (Electronic)	9781538616451
DOIs	https://doi.org/10.1109/SMC.2017.8123045
Publication status	Published - 2017 Nov 27
Event	2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017 - Banff, Canada Duration: 2017 Oct 5 → 2017 Oct 8

Publication series

Name	2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017
Volume	2017-January

Other

Other	2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017
Country/Territory	Canada
City	Banff
Period	17/10/5 → 17/10/8

Bibliographical note

Funding Information:
This study was supported by Basic Science Research Program through the National Research Foundation of KoThis study was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (NRF-2017R1A2B4007970). (Corresponding author: K. Kang; J. Hur; J. Ryu)

Publisher Copyright:
© 2017 IEEE.

ASJC Scopus subject areas

Artificial Intelligence
Computer Science Applications
Human-Computer Interaction
Control and Optimization

Access to Document

10.1109/SMC.2017.8123045

Cite this

Won, J., Kwon, O., Ryu, J., Hur, J., Lee, I., & Kang, K. (2017). TrapFetch: A breakpoint-based prefetcher for both launch and run-time. In 2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017 (pp. 2766-2771). (2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017; Vol. 2017-January). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SMC.2017.8123045

TrapFetch: A breakpoint-based prefetcher for both launch and run-time. / Won, Jiwoong; Kwon, Oseok; Ryu, Junhee et al.
2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017. Institute of Electrical and Electronics Engineers Inc., 2017. p. 2766-2771 (2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017; Vol. 2017-January).

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

Won, J, Kwon, O, Ryu, J, Hur, J, Lee, I & Kang, K 2017, TrapFetch: A breakpoint-based prefetcher for both launch and run-time. in 2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017. 2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017, vol. 2017-January, Institute of Electrical and Electronics Engineers Inc., pp. 2766-2771, 2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017, Banff, Canada, 17/10/5. https://doi.org/10.1109/SMC.2017.8123045

Won J, Kwon O, Ryu J, Hur J, Lee I, Kang K. TrapFetch: A breakpoint-based prefetcher for both launch and run-time. In 2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017. Institute of Electrical and Electronics Engineers Inc. 2017. p. 2766-2771. (2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017). doi: 10.1109/SMC.2017.8123045

@inproceedings{130a251245f347d58493bc0683dcea37,

title = "TrapFetch: A breakpoint-based prefetcher for both launch and run-time",

abstract = "TrapFetch is trained by monitoring the read requests issued by an application. It detects bursts of disk reads, determines the appropriate addresses at which breakpoints should be inserted in the application and library codes prior to the bursts of reads, and then logs this information with the data requested during the interval between each consecutive pair of breakpoints. When the application and library codes are loaded from the disk into memory, TrapFetch inserts breakpoints at the designated addresses based on the logs. During subsequent runs, TrapFetch is invoked at each breakpoint when it prefetches the corresponding data into the page cache. This approach is effective during both launch and run-time. TrapFetch operates at the user level, thus avoiding interference with other applications. In experiments on five popular applications (FlightGear, SpeedDreams 2, Pillars of Eternity, Eclipse, and VegaStrike), TrapFetch reduced the time for launch by up to 39.7% and time for run-time dataloading by up to 63.7%.",

author = "Jiwoong Won and Oseok Kwon and Junhee Ryu and Junbeom Hur and Insup Lee and Kyungtae Kang",

note = "Funding Information: This study was supported by Basic Science Research Program through the National Research Foundation of KoThis study was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (NRF-2017R1A2B4007970). (Corresponding author: K. Kang; J. Hur; J. Ryu) Publisher Copyright: {\textcopyright} 2017 IEEE.; 2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017 ; Conference date: 05-10-2017 Through 08-10-2017",

year = "2017",

month = nov,

day = "27",

doi = "10.1109/SMC.2017.8123045",

language = "English",

series = "2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "2766--2771",

booktitle = "2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017",

}

TY - GEN

T1 - TrapFetch

T2 - 2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017

AU - Won, Jiwoong

AU - Kwon, Oseok

AU - Ryu, Junhee

AU - Hur, Junbeom

AU - Lee, Insup

AU - Kang, Kyungtae

N1 - Funding Information: This study was supported by Basic Science Research Program through the National Research Foundation of KoThis study was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (NRF-2017R1A2B4007970). (Corresponding author: K. Kang; J. Hur; J. Ryu) Publisher Copyright: © 2017 IEEE.

PY - 2017/11/27

Y1 - 2017/11/27

N2 - TrapFetch is trained by monitoring the read requests issued by an application. It detects bursts of disk reads, determines the appropriate addresses at which breakpoints should be inserted in the application and library codes prior to the bursts of reads, and then logs this information with the data requested during the interval between each consecutive pair of breakpoints. When the application and library codes are loaded from the disk into memory, TrapFetch inserts breakpoints at the designated addresses based on the logs. During subsequent runs, TrapFetch is invoked at each breakpoint when it prefetches the corresponding data into the page cache. This approach is effective during both launch and run-time. TrapFetch operates at the user level, thus avoiding interference with other applications. In experiments on five popular applications (FlightGear, SpeedDreams 2, Pillars of Eternity, Eclipse, and VegaStrike), TrapFetch reduced the time for launch by up to 39.7% and time for run-time dataloading by up to 63.7%.

AB - TrapFetch is trained by monitoring the read requests issued by an application. It detects bursts of disk reads, determines the appropriate addresses at which breakpoints should be inserted in the application and library codes prior to the bursts of reads, and then logs this information with the data requested during the interval between each consecutive pair of breakpoints. When the application and library codes are loaded from the disk into memory, TrapFetch inserts breakpoints at the designated addresses based on the logs. During subsequent runs, TrapFetch is invoked at each breakpoint when it prefetches the corresponding data into the page cache. This approach is effective during both launch and run-time. TrapFetch operates at the user level, thus avoiding interference with other applications. In experiments on five popular applications (FlightGear, SpeedDreams 2, Pillars of Eternity, Eclipse, and VegaStrike), TrapFetch reduced the time for launch by up to 39.7% and time for run-time dataloading by up to 63.7%.

UR - http://www.scopus.com/inward/record.url?scp=85044207038&partnerID=8YFLogxK

U2 - 10.1109/SMC.2017.8123045

DO - 10.1109/SMC.2017.8123045

M3 - Conference contribution

AN - SCOPUS:85044207038

T3 - 2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017

SP - 2766

EP - 2771

BT - 2017 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2017

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 5 October 2017 through 8 October 2017

ER -

TrapFetch: A breakpoint-based prefetcher for both launch and run-time

Abstract

Publication series

Other

Bibliographical note

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this