Consecutive job submission behavior at Mira supercomputer

Stephan Schlagkamp; Rafael Ferreira Da Silva; William Allcock; Ewa Deelman; Uwe Schwiegelshohn

doi:10.1145/2907294.2907314

Consecutive job submission behavior at Mira supercomputer

Stephan Schlagkamp
, Rafael Ferreira Da Silva
, William Allcock
, Ewa Deelman
, Uwe Schwiegelshohn

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

21 Scopus citations

Abstract

Understanding user behavior is crucial for the evaluation of scheduling and allocation performances in HPC environments. This paper aims to further understand the dynamic user reaction to different levels of system performance by performing a comprehensive analysis of user behavior in recorded data in the form of delays in the subsequent job submission behavior. Therefore, we characterize a workload trace covering one year of job submissions from the Mira supercomputer at ALCF (Argonne Leadership Computing Facility). We perform an in-depth analysis of correlations between job characteristics, system performance metrics, and the subsequent user behavior. Analysis results show that the user behavior is significantly inuenced by long waiting times, and that complex jobs (number of nodes and CPU hours) lead to longer delays in subsequent job submissions.

Original language	English
Title of host publication	HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing
Publisher	Association for Computing Machinery, Inc
Pages	93-96
Number of pages	4
ISBN (Electronic)	9781450343145
DOIs	https://doi.org/10.1145/2907294.2907314
State	Published - May 31 2016
Externally published	Yes
Event	25th ACM International Symposium on High-Performance Parallel and Distributed Computing, HPDC 2016 - Kyoto, Japan Duration: May 31 2016 → Jun 4 2016

Publication series

Name	HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing

Conference

Conference	25th ACM International Symposium on High-Performance Parallel and Distributed Computing, HPDC 2016
Country/Territory	Japan
City	Kyoto
Period	05/31/16 → 06/4/16

Funding

This work was partly funded by DOE contract number ER26110, "dV/dt - Accelerating the Rate of Progress Towards Extreme Scale Collaborative Science", and #DESC0012636, "Panorama - Predictive Modeling and Diagnostic Monitoring of Extreme Science Workflows", and by the German Research Foundation, RTG 1855. This research used resources of the Argonne Leadership Computing Facility, which is a DOE Office of Science User Facility supported under Contract DE-AC02-06CH11357.

Keywords

Performance modeling
User behavior
Workload analysis

Access to Document

10.1145/2907294.2907314

Cite this

Schlagkamp, S., Da Silva, R. F., Allcock, W., Deelman, E., & Schwiegelshohn, U. (2016). Consecutive job submission behavior at Mira supercomputer. In HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing (pp. 93-96). (HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing). Association for Computing Machinery, Inc. https://doi.org/10.1145/2907294.2907314

Schlagkamp, Stephan ; Da Silva, Rafael Ferreira ; Allcock, William et al. / Consecutive job submission behavior at Mira supercomputer. HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing. Association for Computing Machinery, Inc, 2016. pp. 93-96 (HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing).

@inproceedings{0d8b9389fd4a4d11acf066409ace595f,

title = "Consecutive job submission behavior at Mira supercomputer",

abstract = "Understanding user behavior is crucial for the evaluation of scheduling and allocation performances in HPC environments. This paper aims to further understand the dynamic user reaction to different levels of system performance by performing a comprehensive analysis of user behavior in recorded data in the form of delays in the subsequent job submission behavior. Therefore, we characterize a workload trace covering one year of job submissions from the Mira supercomputer at ALCF (Argonne Leadership Computing Facility). We perform an in-depth analysis of correlations between job characteristics, system performance metrics, and the subsequent user behavior. Analysis results show that the user behavior is significantly inuenced by long waiting times, and that complex jobs (number of nodes and CPU hours) lead to longer delays in subsequent job submissions.",

keywords = "Performance modeling, User behavior, Workload analysis",

author = "Stephan Schlagkamp and \{Da Silva\}, \{Rafael Ferreira\} and William Allcock and Ewa Deelman and Uwe Schwiegelshohn",

note = "Publisher Copyright: Copyright {\textcopyright} 2016 by the Association for Computing Machinery, Inc. (ACM).; 25th ACM International Symposium on High-Performance Parallel and Distributed Computing, HPDC 2016 ; Conference date: 31-05-2016 Through 04-06-2016",

year = "2016",

month = may,

day = "31",

doi = "10.1145/2907294.2907314",

language = "English",

series = "HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing",

publisher = "Association for Computing Machinery, Inc",

pages = "93--96",

booktitle = "HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing",

}

Schlagkamp, S, Da Silva, RF, Allcock, W, Deelman, E & Schwiegelshohn, U 2016, Consecutive job submission behavior at Mira supercomputer. in HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing. HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing, Association for Computing Machinery, Inc, pp. 93-96, 25th ACM International Symposium on High-Performance Parallel and Distributed Computing, HPDC 2016, Kyoto, Japan, 05/31/16. https://doi.org/10.1145/2907294.2907314

Consecutive job submission behavior at Mira supercomputer. / Schlagkamp, Stephan; Da Silva, Rafael Ferreira; Allcock, William et al.
HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing. Association for Computing Machinery, Inc, 2016. p. 93-96 (HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing).

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

TY - GEN

T1 - Consecutive job submission behavior at Mira supercomputer

AU - Schlagkamp, Stephan

AU - Da Silva, Rafael Ferreira

AU - Allcock, William

AU - Deelman, Ewa

AU - Schwiegelshohn, Uwe

PY - 2016/5/31

Y1 - 2016/5/31

N2 - Understanding user behavior is crucial for the evaluation of scheduling and allocation performances in HPC environments. This paper aims to further understand the dynamic user reaction to different levels of system performance by performing a comprehensive analysis of user behavior in recorded data in the form of delays in the subsequent job submission behavior. Therefore, we characterize a workload trace covering one year of job submissions from the Mira supercomputer at ALCF (Argonne Leadership Computing Facility). We perform an in-depth analysis of correlations between job characteristics, system performance metrics, and the subsequent user behavior. Analysis results show that the user behavior is significantly inuenced by long waiting times, and that complex jobs (number of nodes and CPU hours) lead to longer delays in subsequent job submissions.

AB - Understanding user behavior is crucial for the evaluation of scheduling and allocation performances in HPC environments. This paper aims to further understand the dynamic user reaction to different levels of system performance by performing a comprehensive analysis of user behavior in recorded data in the form of delays in the subsequent job submission behavior. Therefore, we characterize a workload trace covering one year of job submissions from the Mira supercomputer at ALCF (Argonne Leadership Computing Facility). We perform an in-depth analysis of correlations between job characteristics, system performance metrics, and the subsequent user behavior. Analysis results show that the user behavior is significantly inuenced by long waiting times, and that complex jobs (number of nodes and CPU hours) lead to longer delays in subsequent job submissions.

KW - Performance modeling

KW - User behavior

KW - Workload analysis

UR - https://www.scopus.com/pages/publications/84978524419

U2 - 10.1145/2907294.2907314

DO - 10.1145/2907294.2907314

M3 - Conference contribution

AN - SCOPUS:84978524419

T3 - HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing

SP - 93

EP - 96

BT - HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing

PB - Association for Computing Machinery, Inc

T2 - 25th ACM International Symposium on High-Performance Parallel and Distributed Computing, HPDC 2016

Y2 - 31 May 2016 through 4 June 2016

ER -

Schlagkamp S, Da Silva RF, Allcock W, Deelman E, Schwiegelshohn U. Consecutive job submission behavior at Mira supercomputer. In HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing. Association for Computing Machinery, Inc. 2016. p. 93-96. (HPDC 2016 - Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing). doi: 10.1145/2907294.2907314

Consecutive job submission behavior at Mira supercomputer

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this