Battle of the defaults: Extracting performance characteristics of HDF5 under production load

Bing Xie; Houjun Tang; Suren Byna; Jesse Hanley; Quincey Koziol; Tonglin Li; Sarp Oral

doi:10.1109/CCGrid51090.2021.00015

Battle of the defaults: Extracting performance characteristics of HDF5 under production load

Bing Xie, Houjun Tang, Suren Byna, Jesse Hanley, Quincey Koziol, Tonglin Li, Sarp Oral

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

7 Scopus citations

Abstract

Popular parallel I/O libraries, such as HDF5, provide tuning parameters to obtain superior performance. However, the selection of effective parameters on production systems is complex due to the interdependence of I/O software and file system layers. Hence, application developers typically use the default parameters and often experience poor I/O performance. This work conducts a benchmarking-based analysis on the HDF5 behaviors with a wide variety of I/O patterns to extract performance characteristics under the production workload. To make the analysis well controlled, we exercise I/O benchmarks on POSIX-IO, MPI-IO, and HDF5 using the same I/O patterns and in the same jobs. To address high performance variability in production environments, we repeat the benchmarks across I/O patterns, storage devices, and time intervals. Based on the results, we identified consistent HDF5 behaviors that appropriate configurations and operations on dataset layout and file-metadata placement can improve performance significantly. We apply our findings and evaluate the tuned I/O library on two supercomputers: Summit and Cori. The results show that our tuned parameters can achieve more than 10× I/O performance speedup than that with default parameters on both systems, suggesting the effectiveness, stability, and generality of our solution.

Original language	English
Title of host publication	Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021
Editors	Laurent Lefevre, Stacy Patterson, Young Choon Lee, Haiying Shen, Shashikant Ilager, Mohammad Goudarzi, Adel N. Toosi, Rajkumar Buyya
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	51-60
Number of pages	10
ISBN (Electronic)	9781728195865
DOIs	https://doi.org/10.1109/CCGrid51090.2021.00015
State	Published - May 2021
Event	21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021 - Virtual, Melbourne, Australia Duration: May 10 2021 → May 13 2021

Publication series

Name	Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021

Conference

Conference	21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021
Country/Territory	Australia
City	Virtual, Melbourne
Period	05/10/21 → 05/13/21

Funding

the U.S. Department of Energy under Contract No. DE-AC02-05CH11231. This work used resources of the Oak Ridge Leadership Computing Facility at the Oak Ridge National Laboratory, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC05-00OR22725 and 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. ACKNOWLEDGMENT This research is supported by the Director, Office of Science, Office of Advanced Scientific Computing Research, of

Access to Document

10.1109/CCGrid51090.2021.00015

Cite this

Xie, B., Tang, H., Byna, S., Hanley, J., Koziol, Q., Li, T., & Oral, S. (2021). Battle of the defaults: Extracting performance characteristics of HDF5 under production load. In L. Lefevre, S. Patterson, Y. C. Lee, H. Shen, S. Ilager, M. Goudarzi, A. N. Toosi, & R. Buyya (Eds.), Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021 (pp. 51-60). (Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CCGrid51090.2021.00015

Xie, Bing ; Tang, Houjun ; Byna, Suren et al. / Battle of the defaults : Extracting performance characteristics of HDF5 under production load. Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021. editor / Laurent Lefevre ; Stacy Patterson ; Young Choon Lee ; Haiying Shen ; Shashikant Ilager ; Mohammad Goudarzi ; Adel N. Toosi ; Rajkumar Buyya. Institute of Electrical and Electronics Engineers Inc., 2021. pp. 51-60 (Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021).

@inproceedings{7584b58a1c194d30b6908bf854687713,

title = "Battle of the defaults: Extracting performance characteristics of HDF5 under production load",

abstract = "Popular parallel I/O libraries, such as HDF5, provide tuning parameters to obtain superior performance. However, the selection of effective parameters on production systems is complex due to the interdependence of I/O software and file system layers. Hence, application developers typically use the default parameters and often experience poor I/O performance. This work conducts a benchmarking-based analysis on the HDF5 behaviors with a wide variety of I/O patterns to extract performance characteristics under the production workload. To make the analysis well controlled, we exercise I/O benchmarks on POSIX-IO, MPI-IO, and HDF5 using the same I/O patterns and in the same jobs. To address high performance variability in production environments, we repeat the benchmarks across I/O patterns, storage devices, and time intervals. Based on the results, we identified consistent HDF5 behaviors that appropriate configurations and operations on dataset layout and file-metadata placement can improve performance significantly. We apply our findings and evaluate the tuned I/O library on two supercomputers: Summit and Cori. The results show that our tuned parameters can achieve more than 10× I/O performance speedup than that with default parameters on both systems, suggesting the effectiveness, stability, and generality of our solution.",

author = "Bing Xie and Houjun Tang and Suren Byna and Jesse Hanley and Quincey Koziol and Tonglin Li and Sarp Oral",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021 ; Conference date: 10-05-2021 Through 13-05-2021",

year = "2021",

month = may,

doi = "10.1109/CCGrid51090.2021.00015",

language = "English",

series = "Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021",

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

pages = "51--60",

editor = "Laurent Lefevre and Stacy Patterson and Lee, {Young Choon} and Haiying Shen and Shashikant Ilager and Mohammad Goudarzi and Toosi, {Adel N.} and Rajkumar Buyya",

booktitle = "Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021",

}

Xie, B, Tang, H, Byna, S, Hanley, J, Koziol, Q, Li, T & Oral, S 2021, Battle of the defaults: Extracting performance characteristics of HDF5 under production load. in L Lefevre, S Patterson, YC Lee, H Shen, S Ilager, M Goudarzi, AN Toosi & R Buyya (eds), Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021. Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021, Institute of Electrical and Electronics Engineers Inc., pp. 51-60, 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021, Virtual, Melbourne, Australia, 05/10/21. https://doi.org/10.1109/CCGrid51090.2021.00015

Battle of the defaults: Extracting performance characteristics of HDF5 under production load. / Xie, Bing; Tang, Houjun; Byna, Suren et al.
Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021. ed. / Laurent Lefevre; Stacy Patterson; Young Choon Lee; Haiying Shen; Shashikant Ilager; Mohammad Goudarzi; Adel N. Toosi; Rajkumar Buyya. Institute of Electrical and Electronics Engineers Inc., 2021. p. 51-60 (Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021).

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

TY - GEN

T1 - Battle of the defaults

T2 - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021

AU - Xie, Bing

AU - Tang, Houjun

AU - Byna, Suren

AU - Hanley, Jesse

AU - Koziol, Quincey

AU - Li, Tonglin

AU - Oral, Sarp

PY - 2021/5

Y1 - 2021/5

N2 - Popular parallel I/O libraries, such as HDF5, provide tuning parameters to obtain superior performance. However, the selection of effective parameters on production systems is complex due to the interdependence of I/O software and file system layers. Hence, application developers typically use the default parameters and often experience poor I/O performance. This work conducts a benchmarking-based analysis on the HDF5 behaviors with a wide variety of I/O patterns to extract performance characteristics under the production workload. To make the analysis well controlled, we exercise I/O benchmarks on POSIX-IO, MPI-IO, and HDF5 using the same I/O patterns and in the same jobs. To address high performance variability in production environments, we repeat the benchmarks across I/O patterns, storage devices, and time intervals. Based on the results, we identified consistent HDF5 behaviors that appropriate configurations and operations on dataset layout and file-metadata placement can improve performance significantly. We apply our findings and evaluate the tuned I/O library on two supercomputers: Summit and Cori. The results show that our tuned parameters can achieve more than 10× I/O performance speedup than that with default parameters on both systems, suggesting the effectiveness, stability, and generality of our solution.

AB - Popular parallel I/O libraries, such as HDF5, provide tuning parameters to obtain superior performance. However, the selection of effective parameters on production systems is complex due to the interdependence of I/O software and file system layers. Hence, application developers typically use the default parameters and often experience poor I/O performance. This work conducts a benchmarking-based analysis on the HDF5 behaviors with a wide variety of I/O patterns to extract performance characteristics under the production workload. To make the analysis well controlled, we exercise I/O benchmarks on POSIX-IO, MPI-IO, and HDF5 using the same I/O patterns and in the same jobs. To address high performance variability in production environments, we repeat the benchmarks across I/O patterns, storage devices, and time intervals. Based on the results, we identified consistent HDF5 behaviors that appropriate configurations and operations on dataset layout and file-metadata placement can improve performance significantly. We apply our findings and evaluate the tuned I/O library on two supercomputers: Summit and Cori. The results show that our tuned parameters can achieve more than 10× I/O performance speedup than that with default parameters on both systems, suggesting the effectiveness, stability, and generality of our solution.

UR - http://www.scopus.com/inward/record.url?scp=85114888893&partnerID=8YFLogxK

U2 - 10.1109/CCGrid51090.2021.00015

DO - 10.1109/CCGrid51090.2021.00015

M3 - Conference contribution

AN - SCOPUS:85114888893

T3 - Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021

SP - 51

EP - 60

BT - Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021

A2 - Lefevre, Laurent

A2 - Patterson, Stacy

A2 - Lee, Young Choon

A2 - Shen, Haiying

A2 - Ilager, Shashikant

A2 - Goudarzi, Mohammad

A2 - Toosi, Adel N.

A2 - Buyya, Rajkumar

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 10 May 2021 through 13 May 2021

ER -

Xie B, Tang H, Byna S, Hanley J, Koziol Q, Li T et al. Battle of the defaults: Extracting performance characteristics of HDF5 under production load. In Lefevre L, Patterson S, Lee YC, Shen H, Ilager S, Goudarzi M, Toosi AN, Buyya R, editors, Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021. Institute of Electrical and Electronics Engineers Inc. 2021. p. 51-60. (Proceedings - 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, CCGrid 2021). doi: 10.1109/CCGrid51090.2021.00015

Battle of the defaults: Extracting performance characteristics of HDF5 under production load

Abstract

Publication series

Conference

Funding

Access to Document

Other files and links

Fingerprint

Cite this