A novel metric to evaluate in situ workflows

Tu Mai Anh Do; Loïc Pottier; Stephen Thomas; Rafael Ferreira da Silva; Michel A. Cuendet; Harel Weinstein; Trilce Estrada; Michela Taufer; Ewa Deelman

doi:10.1007/978-3-030-50371-0_40

A novel metric to evaluate in situ workflows

Tu Mai Anh Do, Loïc Pottier, Stephen Thomas, Rafael Ferreira da Silva, Michel A. Cuendet, Harel Weinstein, Trilce Estrada, Michela Taufer, Ewa Deelman

Workflow and Ecosystem Services

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

6 Scopus citations

Abstract

Performance evaluation is crucial to understanding the behavior of scientific workflows and efficiently utilizing resources on high-performance computing architectures. In this study, we target an emerging type of workflow, called in situ workflows. Through an analysis of the state-of-the-art research on in situ workflows, we model a theoretical framework that helps characterize such workflows. We further propose a lightweight metric for assessing resource usage efficiency of an in situ workflow execution. By applying this metric to a simple, yet representative, synthetic workflow, we explore two possible scenarios (Idle Simulation and Idle Analyzer) for the execution of real in situ workflows. Experimental results show that there is no substantial difference in the performance of both the in transit placement (analytics on dedicated nodes) and the helper-core configuration (analytics co-allocated with simulation) on our target system.

Original language	English
Title of host publication	Computational Science – ICCS 2020 - 20th International Conference, Proceedings
Editors	Valeria V. Krzhizhanovskaya, Gábor Závodszky, Michael H. Lees, Peter M.A. Sloot, Peter M.A. Sloot, Peter M.A. Sloot, Jack J. Dongarra, Sérgio Brissos, João Teixeira
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	538-553
Number of pages	16
ISBN (Print)	9783030503703
DOIs	https://doi.org/10.1007/978-3-030-50371-0_40
State	Published - 2020
Event	20th International Conference on Computational Science, ICCS 2020 - Amsterdam, Netherlands Duration: Jun 3 2020 → Jun 5 2020

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12137 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	20th International Conference on Computational Science, ICCS 2020
Country/Territory	Netherlands
City	Amsterdam
Period	06/3/20 → 06/5/20

Funding

This work is funded by NSF contracts #1741040, #1740990 and #1741057; and DOE contract #DE-SC0012636. We are grateful to IBM for the Shared University Research Award that supported the purchase of IBM Power9 system used in this paper.

Keywords

High-performance computing
In situ model
Molecular dynamics
Scientific workflow

Access to Document

10.1007/978-3-030-50371-0_40

Cite this

Do, T. M. A., Pottier, L., Thomas, S., da Silva, R. F., Cuendet, M. A., Weinstein, H., Estrada, T., Taufer, M., & Deelman, E. (2020). A novel metric to evaluate in situ workflows. In V. V. Krzhizhanovskaya, G. Závodszky, M. H. Lees, P. M. A. Sloot, P. M. A. Sloot, P. M. A. Sloot, J. J. Dongarra, S. Brissos, & J. Teixeira (Eds.), Computational Science – ICCS 2020 - 20th International Conference, Proceedings (pp. 538-553). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12137 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-50371-0_40

Do, Tu Mai Anh ; Pottier, Loïc ; Thomas, Stephen et al. / A novel metric to evaluate in situ workflows. Computational Science – ICCS 2020 - 20th International Conference, Proceedings. editor / Valeria V. Krzhizhanovskaya ; Gábor Závodszky ; Michael H. Lees ; Peter M.A. Sloot ; Peter M.A. Sloot ; Peter M.A. Sloot ; Jack J. Dongarra ; Sérgio Brissos ; João Teixeira. Springer Science and Business Media Deutschland GmbH, 2020. pp. 538-553 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{5611e07efedb408a94df92c4d0172f3f,

title = "A novel metric to evaluate in situ workflows",

abstract = "Performance evaluation is crucial to understanding the behavior of scientific workflows and efficiently utilizing resources on high-performance computing architectures. In this study, we target an emerging type of workflow, called in situ workflows. Through an analysis of the state-of-the-art research on in situ workflows, we model a theoretical framework that helps characterize such workflows. We further propose a lightweight metric for assessing resource usage efficiency of an in situ workflow execution. By applying this metric to a simple, yet representative, synthetic workflow, we explore two possible scenarios (Idle Simulation and Idle Analyzer) for the execution of real in situ workflows. Experimental results show that there is no substantial difference in the performance of both the in transit placement (analytics on dedicated nodes) and the helper-core configuration (analytics co-allocated with simulation) on our target system.",

keywords = "High-performance computing, In situ model, Molecular dynamics, Scientific workflow",

author = "Do, {Tu Mai Anh} and Lo{\"i}c Pottier and Stephen Thomas and {da Silva}, {Rafael Ferreira} and Cuendet, {Michel A.} and Harel Weinstein and Trilce Estrada and Michela Taufer and Ewa Deelman",

note = "Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2020.; 20th International Conference on Computational Science, ICCS 2020 ; Conference date: 03-06-2020 Through 05-06-2020",

year = "2020",

doi = "10.1007/978-3-030-50371-0_40",

language = "English",

isbn = "9783030503703",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "538--553",

editor = "Krzhizhanovskaya, {Valeria V.} and G{\'a}bor Z{\'a}vodszky and Lees, {Michael H.} and Sloot, {Peter M.A.} and Sloot, {Peter M.A.} and Sloot, {Peter M.A.} and Dongarra, {Jack J.} and S{\'e}rgio Brissos and Jo{\~a}o Teixeira",

booktitle = "Computational Science – ICCS 2020 - 20th International Conference, Proceedings",

}

Do, TMA, Pottier, L, Thomas, S, da Silva, RF, Cuendet, MA, Weinstein, H, Estrada, T, Taufer, M & Deelman, E 2020, A novel metric to evaluate in situ workflows. in VV Krzhizhanovskaya, G Závodszky, MH Lees, PMA Sloot, PMA Sloot, PMA Sloot, JJ Dongarra, S Brissos & J Teixeira (eds), Computational Science – ICCS 2020 - 20th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12137 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 538-553, 20th International Conference on Computational Science, ICCS 2020, Amsterdam, Netherlands, 06/3/20. https://doi.org/10.1007/978-3-030-50371-0_40

A novel metric to evaluate in situ workflows. / Do, Tu Mai Anh; Pottier, Loïc; Thomas, Stephen et al.
Computational Science – ICCS 2020 - 20th International Conference, Proceedings. ed. / Valeria V. Krzhizhanovskaya; Gábor Závodszky; Michael H. Lees; Peter M.A. Sloot; Peter M.A. Sloot; Peter M.A. Sloot; Jack J. Dongarra; Sérgio Brissos; João Teixeira. Springer Science and Business Media Deutschland GmbH, 2020. p. 538-553 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12137 LNCS).

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

TY - GEN

T1 - A novel metric to evaluate in situ workflows

AU - Do, Tu Mai Anh

AU - Pottier, Loïc

AU - Thomas, Stephen

AU - da Silva, Rafael Ferreira

AU - Cuendet, Michel A.

AU - Weinstein, Harel

AU - Estrada, Trilce

AU - Taufer, Michela

AU - Deelman, Ewa

N1 - Publisher Copyright: © Springer Nature Switzerland AG 2020.

PY - 2020

Y1 - 2020

N2 - Performance evaluation is crucial to understanding the behavior of scientific workflows and efficiently utilizing resources on high-performance computing architectures. In this study, we target an emerging type of workflow, called in situ workflows. Through an analysis of the state-of-the-art research on in situ workflows, we model a theoretical framework that helps characterize such workflows. We further propose a lightweight metric for assessing resource usage efficiency of an in situ workflow execution. By applying this metric to a simple, yet representative, synthetic workflow, we explore two possible scenarios (Idle Simulation and Idle Analyzer) for the execution of real in situ workflows. Experimental results show that there is no substantial difference in the performance of both the in transit placement (analytics on dedicated nodes) and the helper-core configuration (analytics co-allocated with simulation) on our target system.

AB - Performance evaluation is crucial to understanding the behavior of scientific workflows and efficiently utilizing resources on high-performance computing architectures. In this study, we target an emerging type of workflow, called in situ workflows. Through an analysis of the state-of-the-art research on in situ workflows, we model a theoretical framework that helps characterize such workflows. We further propose a lightweight metric for assessing resource usage efficiency of an in situ workflow execution. By applying this metric to a simple, yet representative, synthetic workflow, we explore two possible scenarios (Idle Simulation and Idle Analyzer) for the execution of real in situ workflows. Experimental results show that there is no substantial difference in the performance of both the in transit placement (analytics on dedicated nodes) and the helper-core configuration (analytics co-allocated with simulation) on our target system.

KW - High-performance computing

KW - In situ model

KW - Molecular dynamics

KW - Scientific workflow

UR - http://www.scopus.com/inward/record.url?scp=85087529021&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-50371-0_40

DO - 10.1007/978-3-030-50371-0_40

M3 - Conference contribution

AN - SCOPUS:85087529021

SN - 9783030503703

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 538

EP - 553

BT - Computational Science – ICCS 2020 - 20th International Conference, Proceedings

A2 - Krzhizhanovskaya, Valeria V.

A2 - Závodszky, Gábor

A2 - Lees, Michael H.

A2 - Sloot, Peter M.A.

A2 - Dongarra, Jack J.

A2 - Brissos, Sérgio

A2 - Teixeira, João

PB - Springer Science and Business Media Deutschland GmbH

T2 - 20th International Conference on Computational Science, ICCS 2020

Y2 - 3 June 2020 through 5 June 2020

ER -

Do TMA, Pottier L, Thomas S, da Silva RF, Cuendet MA, Weinstein H et al. A novel metric to evaluate in situ workflows. In Krzhizhanovskaya VV, Závodszky G, Lees MH, Sloot PMA, Sloot PMA, Sloot PMA, Dongarra JJ, Brissos S, Teixeira J, editors, Computational Science – ICCS 2020 - 20th International Conference, Proceedings. Springer Science and Business Media Deutschland GmbH. 2020. p. 538-553. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-50371-0_40

A novel metric to evaluate in situ workflows

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this