Exploring Large All-Flash Storage System with Scientific Simulation
Authors: 
Junmin Gu, 
Greg Eisenhauer, Scott Klasky, Norbert Podhorszki, Ruonan Wang, 
Kesheng WuAuthors Info & Claims
Article No.: 23, Pages 1 - 4
Abstract
Solid state storage systems have been very effectively used in small devices; however, their effectiveness for large systems such as supercomputers is not yet proven. Recently, for the first time, a new supercomputer is being deployed with an all-flash storage as its main file system. In this work, we report our preliminary study of the I/O performance on this supercomputer named Perlmutter. We are able to achieve 1.4 TB/s with the default file configuration on the system. This default configuration outperforms dozens of other choices tested, though the current observed performance is still pretty far from the theoretical peak performance of 5 TB/s.
References
[1]
Peter Braam. 2019. The Lustre storage architecture. Cluster File Systems, Inc. https://doi.org/10.48550/arXiv.1903.01955 arXiv preprint arXiv:1903.01955.
[2]
[2] NERSC documentation.2021. https://docs.nersc.gov/performance/io/lustre/.
[3]
NERSC documentation. 2021. Nersc Systems Nersc file system Description. NERSC. https://docs.nersc.gov/systems/
[4]
William F. Godoy, Norbert Podhorszki, Ruonan Wang, Chuck Atkins, Greg Eisenhauer, Junmin Gu, Philip Davis, Jong Choi, Kai Germaschewski, Kevin Huck, Axel Huebl, Mark Kim, James Kress, Tahsin Kurc, Qing Liu, Jeremy Logan, Kshitij Mehta, George Ostrouchov, Manish Parashar, Franz Poeschel, David Pugmire, Eric Suchyta, Keichi Takahashi, Nick Thompson, Seiji Tsutsumi, Lipeng Wan, Matthew Wolf, Kesheng Wu, and Scott Klasky. 2020. ADIOS2: The Adaptable Input Output System. A framework for high-performance data management. SoftwareX 12(2020), 100561. https://doi.org/10.1016/j.softx.2020.100561
[5]
Junmin Gu, Philip Davis, Greg Eisenhauer, William Godoy, Axel Huebl, Scott Klasky, Manish Parashar, Norbert Podhorszki, Franz Poeschel, JeanLuc Vay, Lipeng Wan, Ruonan Wang, and Kesheng Wu. 2022. Organizing Large Data Sets for Efficient Analyses on HPC Systems. Journal of Physics: Conference Series 2224, 1 (2022), 012042. https://doi.org/10.1088/1742-6596/2224/1/012042
[6]
Junmin Gu, Scott Klasky, Norbert Podhorszki, Ji Qiang, and Kesheng Wu. 2018. Querying large scientific data sets with adaptable IO system ADIOS. In Asian Conference on Supercomputing Frontiers. Springer, Springer, 51–69.
[7]
T. Hey, S. Tansley, and K. Tolle (Eds.). 2009. The Fourth Paradigm: Data-Intensive Scientific Discovery. Microsoft. 287 pages.
[8]
Sunggon Kim, Alex Sim, Kesheng Wu, Suren Byna, Yongseok Son, and Hyeonsang Eom. 2020. Towards HPC I/O performance prediction through large-scale log analysis. In Proceedings of the 29th International Symposium on High-Performance Parallel and Distributed Computing. ACM, 77–88. https://doi.org/10.1145/3369583.3392678
[9]
Fabian Koller, Franz Poeschel, Junmin Gu, and Axel Huebl. 2021. openPMD-api: C++ & Python API for Scientific I/O with openPMD. OpenPMD. https://doi.org/10.14278/rodare.27 https://github.com/openPMD/openPMD-api.
[10]
Donghun Koo, Jaehwan Lee, Jialin Liu, Eun-Kyu Byun, Jae-Hyuck Kwak, Glenn K. Lockwood, Soonwook Hwang, Katie Antypas, Kesheng Wu, and Hyeonsang Eom. 2021. An empirical study of I/O separation for burst buffers in HPC systems. J. Parallel and Distrib. Comput. 148 (2021), 96–108. https://doi.org/10.1016/j.jpdc.2020.10.007
[11]
Adam Leventhal. 2008. Flash Storage Today: Can Flash Memory Become the Foundation for a New Tier in the Storage Hierarchy?Queue 6, 4 (jul 2008), 24––30. https://doi.org/10.1145/1413254.1413262
[12]
Glenn K. Lockwood, Alberto Chiusole, and Nicholas J. Wright. 2021. New Challenges of Benchmarking All-Flash Storage for HPC. In 2021 IEEE/ACM Sixth International Parallel Data Systems Workshop (PDSW). IEEE, 1–8. https://doi.org/10.1109/PDSW54622.2021.00006
[13]
Simone Ferlin Oliveira, Karl Furlinger, and Dieter Kranzlmuller. 2012. Trends in Computation, Communication and Storage and the Consequences for Data-intensive Science. In 2012 IEEE 14th International Conference on High Performance Computing and Communication. 572–579. https://doi.org/10.1109/HPCC.2012.83
[14]
Wei Pan, Zhanhuai Li, Yansong Zhang, and Chuliang Weng. 2018. The new hardware development trend and the challenges in data management and analysis. Data Science and Engineering 3, 3 (2018), 263–276. https://doi.org/10.1007/s41019-018-0072-6
[15]
H. Riggs, S. Tufail, I. Parvez, and A. Sarwat. 2020. Survey of Solid State Drives, Characteristics, Technology, and Applications. In 2020 SoutheastCon. IEEE, 1–6. https://doi.org/10.1109/SoutheastCon44009.2020.9249760
[16]
Hanul Sung, Jiwoo Bang, Alexander Sim, Kesheng Wu, and Hyeonsang Eom. 2019. Understanding Parallel I/O Performance Trends Under Various HPC Configurations. In SNTA@HPDC 2019. ACM, 29–36. https://doi.org/10.1145/3322798.3329258
[17]
J.-L. Vay, A. Huebl, A. Almgren, L. D. Amorim, J. Bell, L. Fedeli, L. Ge, K. Gott, D. P. Grote, M. Hogan, R. Jambunathan, R. Lehe, A. Myers, C. Ng, M. Rowan, O. Shapoval, M. Thevenet, H. Vincenti, E. Yang, N. Zaïm, W. Zhang, Y. Zhao, and E. Zoni. 2021. Modeling of a chain of three plasma accelerator stages with the WarpX electromagnetic PIC code on GPUs. Physics of Plasmas 28(2021), 023105. https://doi.org/10.1063/5.0028512
Recommendations
Hybrid hierarchy storage system in MilkyWay-2 supercomputer
With the rapid improvement of computation capability in high performance supercomputer system, the imbalance of performance between computation subsystem and storage subsystem has become more and more serious, especially when various big data are ...
Exploring large-scale small file storage for search engines
Large-scale small file storage for original pages degrades performance of search engines. In this paper, we first analyze the disadvantages of the existing EXT3 file system in accessing small files. Then, the rate and speed of compression algorithms are ...
Comments
Please enable JavaScript to view thecomments powered by Disqus.Information & Contributors
Information
Published In
July 2022
201 pages
Copyright © 2022 Owner/Author.
This work is licensed under a Creative Commons Attribution International 4.0 License.
Publisher
Association for Computing Machinery
New York, NY, United States
Publication History
Published: 23 August 2022
Check for updates
Author Tags
Qualifiers
- Short-paper
- Research
- Refereed limited
Funding Sources
- US Department of Energy
Conference
SSDBM 2022
SSDBM 2022: 34th International Conference on Scientific and Statistical Database Management
July 6 - 8, 2022
Copenhagen, Denmark
Acceptance Rates
Overall Acceptance Rate 56 of 146 submissions, 38%
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 361Total Downloads
- Downloads (Last 12 months)164
- Downloads (Last 6 weeks)15
Reflects downloads up to 02 Oct 2024
Other Metrics
Citations
View Options
View options
View or Download as a PDF file.
PDFeReader
View online with eReader.
eReaderHTML Format
View this article in HTML Format.
HTML FormatGet Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in