On the use of burst buffers for accelerating data-intensive scientific work flows

Rafael Ferreira Da Silva; Scott Callaghan; Ewa Deelman

doi:10.1145/3150994.3151000

On the use of burst buffers for accelerating data-intensive scientific work flows

Rafael Ferreira Da Silva
, Scott Callaghan
, Ewa Deelman

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

12 Scopus citations

Abstract

Science applications frequently produce and consume large volumes of data, but delivering this data to and from compute resources can be challenging, as parallel file system performance is not keeping up with compute and memory performance. To mitigate this I/O bottleneck, some systems have deployed burst buffers, but their impact on performance for real-world workflow applications is not always clear. In this paper, we examine the impact of burst buffers through the remote-shared, allocatable burst buffers on the Cori system at NERSC. By running a subset of the SCEC CyberShake workflow, a production seismic hazard analysis workflow, we find that using burst buffers offers read and write improvements of about an order of magnitude, and these improvements lead to increased job performance, even for long-running CPU-bound jobs.

Original language	English
Title of host publication	Proceedings of WORKS 2017
Subtitle of host publication	12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis
Publisher	Association for Computing Machinery, Inc
ISBN (Print)	9781450351294
DOIs	https://doi.org/10.1145/3150994.3151000
State	Published - Nov 12 2017
Externally published	Yes
Event	12th Workshop on Workflows in Support of Large-Scale Science, WORKS 2017 - Held in conjunction with the International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2017 - Denver, United States Duration: Nov 12 2017 → Nov 17 2017

Publication series

Name	Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis

Conference

Conference	12th Workshop on Workflows in Support of Large-Scale Science, WORKS 2017 - Held in conjunction with the International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2017
Country/Territory	United States
City	Denver
Period	11/12/17 → 11/17/17

Funding

CyberShake workflow research was supported by the National Science Foundation (NSF) under the OAC SI2-SSI grant #1148493, the OAC SI2-SSI grant #1450451, and EAR grant #1226343. This research was supported by the Southern California Earthquake Center (Contribution No. 7610). SCEC is funded by NSF Cooperative Agreement EAR-1033462 & USGS Cooperative Agreement G12AC20038. This work was funded by DOE contract number #DESC0012636, “Panorama – Predictive Modeling and Diagnostic Monitoring of Extreme Science Workflows”, and by NSF contract number #1664162, “ SI2-SSI: Pegasus: Automating Compute and Data Intensive Science”. This research used resources of the National Energy Research Scientific Computing Center, a DOE Office of Science User Facility supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231.

Keywords

Burst Buffers
High-Performance Computing
In Transit Processing
Scientific Workflows

Access to Document

10.1145/3150994.3151000

Cite this

Da Silva, R. F., Callaghan, S., & Deelman, E. (2017). On the use of burst buffers for accelerating data-intensive scientific work flows. In Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis Article 2 (Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis). Association for Computing Machinery, Inc. https://doi.org/10.1145/3150994.3151000

Da Silva, Rafael Ferreira ; Callaghan, Scott ; Deelman, Ewa. / On the use of burst buffers for accelerating data-intensive scientific work flows. Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis. Association for Computing Machinery, Inc, 2017. (Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis).

@inproceedings{5af17e9aa8cd4668b8fb3356105c5093,

title = "On the use of burst buffers for accelerating data-intensive scientific work flows",

abstract = "Science applications frequently produce and consume large volumes of data, but delivering this data to and from compute resources can be challenging, as parallel file system performance is not keeping up with compute and memory performance. To mitigate this I/O bottleneck, some systems have deployed burst buffers, but their impact on performance for real-world workflow applications is not always clear. In this paper, we examine the impact of burst buffers through the remote-shared, allocatable burst buffers on the Cori system at NERSC. By running a subset of the SCEC CyberShake workflow, a production seismic hazard analysis workflow, we find that using burst buffers offers read and write improvements of about an order of magnitude, and these improvements lead to increased job performance, even for long-running CPU-bound jobs.",

keywords = "Burst Buffers, High-Performance Computing, In Transit Processing, Scientific Workflows",

author = "\{Da Silva\}, \{Rafael Ferreira\} and Scott Callaghan and Ewa Deelman",

note = "Publisher Copyright: {\textcopyright} 2017 Association for Computing Machinery.; 12th Workshop on Workflows in Support of Large-Scale Science, WORKS 2017 - Held in conjunction with the International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2017 ; Conference date: 12-11-2017 Through 17-11-2017",

year = "2017",

month = nov,

day = "12",

doi = "10.1145/3150994.3151000",

language = "English",

isbn = "9781450351294",

series = "Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis",

publisher = "Association for Computing Machinery, Inc",

booktitle = "Proceedings of WORKS 2017",

}

Da Silva, RF, Callaghan, S & Deelman, E 2017, On the use of burst buffers for accelerating data-intensive scientific work flows. in Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis., 2, Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis, Association for Computing Machinery, Inc, 12th Workshop on Workflows in Support of Large-Scale Science, WORKS 2017 - Held in conjunction with the International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2017, Denver, United States, 11/12/17. https://doi.org/10.1145/3150994.3151000

On the use of burst buffers for accelerating data-intensive scientific work flows. / Da Silva, Rafael Ferreira; Callaghan, Scott; Deelman, Ewa.
Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis. Association for Computing Machinery, Inc, 2017. 2 (Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis).

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

TY - GEN

T1 - On the use of burst buffers for accelerating data-intensive scientific work flows

AU - Da Silva, Rafael Ferreira

AU - Callaghan, Scott

AU - Deelman, Ewa

PY - 2017/11/12

Y1 - 2017/11/12

N2 - Science applications frequently produce and consume large volumes of data, but delivering this data to and from compute resources can be challenging, as parallel file system performance is not keeping up with compute and memory performance. To mitigate this I/O bottleneck, some systems have deployed burst buffers, but their impact on performance for real-world workflow applications is not always clear. In this paper, we examine the impact of burst buffers through the remote-shared, allocatable burst buffers on the Cori system at NERSC. By running a subset of the SCEC CyberShake workflow, a production seismic hazard analysis workflow, we find that using burst buffers offers read and write improvements of about an order of magnitude, and these improvements lead to increased job performance, even for long-running CPU-bound jobs.

AB - Science applications frequently produce and consume large volumes of data, but delivering this data to and from compute resources can be challenging, as parallel file system performance is not keeping up with compute and memory performance. To mitigate this I/O bottleneck, some systems have deployed burst buffers, but their impact on performance for real-world workflow applications is not always clear. In this paper, we examine the impact of burst buffers through the remote-shared, allocatable burst buffers on the Cori system at NERSC. By running a subset of the SCEC CyberShake workflow, a production seismic hazard analysis workflow, we find that using burst buffers offers read and write improvements of about an order of magnitude, and these improvements lead to increased job performance, even for long-running CPU-bound jobs.

KW - Burst Buffers

KW - High-Performance Computing

KW - In Transit Processing

KW - Scientific Workflows

UR - https://www.scopus.com/pages/publications/85054719710

U2 - 10.1145/3150994.3151000

DO - 10.1145/3150994.3151000

M3 - Conference contribution

AN - SCOPUS:85054719710

SN - 9781450351294

T3 - Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis

BT - Proceedings of WORKS 2017

PB - Association for Computing Machinery, Inc

T2 - 12th Workshop on Workflows in Support of Large-Scale Science, WORKS 2017 - Held in conjunction with the International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2017

Y2 - 12 November 2017 through 17 November 2017

ER -

Da Silva RF, Callaghan S, Deelman E. On the use of burst buffers for accelerating data-intensive scientific work flows. In Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis. Association for Computing Machinery, Inc. 2017. 2. (Proceedings of WORKS 2017: 12th Workshop on Workflows in Support of Large-Scale Science - Held in conjunction with SC 2017: The International Conference for High Performance Computing, Networking, Storage and Analysis). doi: 10.1145/3150994.3151000

On the use of burst buffers for accelerating data-intensive scientific work flows

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this