Implementation and usage of the PERUSE-interface in open MPI

Rainer Keller; George Bosilca; Graham Fagg; Michael Resch; Jack J. Dongarra

doi:10.1007/11846802_48

Implementation and usage of the PERUSE-interface in open MPI

Rainer Keller, George Bosilca, Graham Fagg, Michael Resch, Jack J. Dongarra

Research output: Chapter in Book/Conference proceeding › Conference contribution

Abstract

This paper describes the implementation, usage and experience with the MPI performance revealing extension interface (Peruse) into the Open MPI implementation. While the PMPI-interface allows timing MPI-functions through wrappers, it can not provide MPI-internal information on MPI-states and lower-level network performance. We introduce the general design criteria of the interface implementation and analyze the overhead generated by this functionality. To support performance evaluation of large-scale applications, tools for visualization are imperative. We extend the tracing library of the Paraver-toolkit to support tracing Peruse-events and show how this helps detecting performance bottlenecks. A test-suite and a real-world application are traced and visualized using Paraver. © Springer-Verlag Berlin Heidelberg 2006.

Original language	English
Title of host publication	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)\|Lect. Notes Comput. Sci.
Publisher	Springer Nature
Pages	347-355
Number of pages	8
Volume	4192
ISBN (Print)	354039110X, 9783540391104
DOIs	https://doi.org/10.1007/11846802_48
Publication status	Published - 2006
Event	13th European PVM/MPI User's Group Meeting - Bonn Duration: 1 Jul 2006 → … http://dblp.uni-trier.de/db/conf/pvm/pvm2006.html#Pjesivac-GrbovicFABD06http://dblp.uni-trier.de/rec/bibtex/conf/pvm/Pjesivac-GrbovicFABD06.xmlhttp://dblp.uni-trier.de/rec/bibtex/conf/pvm/Pjesivac-GrbovicFABD06

Publication series

Name	Lecture Notes in Computer Science

Conference

Conference	13th European PVM/MPI User's Group Meeting
City	Bonn
Period	1/07/06 → …
Internet address	http://dblp.uni-trier.de/db/conf/pvm/pvm2006.html#Pjesivac-GrbovicFABD06http://dblp.uni-trier.de/rec/bibtex/conf/pvm/Pjesivac-GrbovicFABD06.xmlhttp://dblp.uni-trier.de/rec/bibtex/conf/pvm/Pjesivac-GrbovicFABD06

Access to Document

10.1007/11846802_48

Cite this

Keller, R., Bosilca, G., Fagg, G., Resch, M., & Dongarra, J. J. (2006). Implementation and usage of the PERUSE-interface in open MPI. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|Lect. Notes Comput. Sci. (Vol. 4192, pp. 347-355). (Lecture Notes in Computer Science). Springer Nature. https://doi.org/10.1007/11846802_48

@inproceedings{a67d49a97af94b1c94983be5d568dedf,

title = "Implementation and usage of the PERUSE-interface in open MPI",

abstract = "This paper describes the implementation, usage and experience with the MPI performance revealing extension interface (Peruse) into the Open MPI implementation. While the PMPI-interface allows timing MPI-functions through wrappers, it can not provide MPI-internal information on MPI-states and lower-level network performance. We introduce the general design criteria of the interface implementation and analyze the overhead generated by this functionality. To support performance evaluation of large-scale applications, tools for visualization are imperative. We extend the tracing library of the Paraver-toolkit to support tracing Peruse-events and show how this helps detecting performance bottlenecks. A test-suite and a real-world application are traced and visualized using Paraver. {\textcopyright} Springer-Verlag Berlin Heidelberg 2006.",

author = "Rainer Keller and George Bosilca and Graham Fagg and Michael Resch and Dongarra, \{Jack J.\}",

year = "2006",

doi = "10.1007/11846802\_48",

language = "English",

isbn = "354039110X",

volume = "4192",

series = "Lecture Notes in Computer Science",

publisher = "Springer Nature",

pages = "347--355",

booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|Lect. Notes Comput. Sci.",

address = "United States",

note = "13th European PVM/MPI User's Group Meeting ; Conference date: 01-07-2006",

url = "http://dblp.uni-trier.de/db/conf/pvm/pvm2006.html\#Pjesivac-GrbovicFABD06http://dblp.uni-trier.de/rec/bibtex/conf/pvm/Pjesivac-GrbovicFABD06.xmlhttp://dblp.uni-trier.de/rec/bibtex/conf/pvm/Pjesivac-GrbovicFABD06",

}

Keller, R, Bosilca, G, Fagg, G, Resch, M & Dongarra, JJ 2006, Implementation and usage of the PERUSE-interface in open MPI. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|Lect. Notes Comput. Sci.. vol. 4192, Lecture Notes in Computer Science, Springer Nature, pp. 347-355, 13th European PVM/MPI User's Group Meeting, Bonn, 1/07/06. https://doi.org/10.1007/11846802_48

Implementation and usage of the PERUSE-interface in open MPI. / Keller, Rainer; Bosilca, George; Fagg, Graham et al.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|Lect. Notes Comput. Sci.. Vol. 4192 Springer Nature, 2006. p. 347-355 (Lecture Notes in Computer Science).

Research output: Chapter in Book/Conference proceeding › Conference contribution

TY - GEN

T1 - Implementation and usage of the PERUSE-interface in open MPI

AU - Keller, Rainer

AU - Bosilca, George

AU - Fagg, Graham

AU - Resch, Michael

AU - Dongarra, Jack J.

PY - 2006

Y1 - 2006

N2 - This paper describes the implementation, usage and experience with the MPI performance revealing extension interface (Peruse) into the Open MPI implementation. While the PMPI-interface allows timing MPI-functions through wrappers, it can not provide MPI-internal information on MPI-states and lower-level network performance. We introduce the general design criteria of the interface implementation and analyze the overhead generated by this functionality. To support performance evaluation of large-scale applications, tools for visualization are imperative. We extend the tracing library of the Paraver-toolkit to support tracing Peruse-events and show how this helps detecting performance bottlenecks. A test-suite and a real-world application are traced and visualized using Paraver. © Springer-Verlag Berlin Heidelberg 2006.

AB - This paper describes the implementation, usage and experience with the MPI performance revealing extension interface (Peruse) into the Open MPI implementation. While the PMPI-interface allows timing MPI-functions through wrappers, it can not provide MPI-internal information on MPI-states and lower-level network performance. We introduce the general design criteria of the interface implementation and analyze the overhead generated by this functionality. To support performance evaluation of large-scale applications, tools for visualization are imperative. We extend the tracing library of the Paraver-toolkit to support tracing Peruse-events and show how this helps detecting performance bottlenecks. A test-suite and a real-world application are traced and visualized using Paraver. © Springer-Verlag Berlin Heidelberg 2006.

UR - https://www.scopus.com/pages/publications/33750226019

U2 - 10.1007/11846802_48

DO - 10.1007/11846802_48

M3 - Conference contribution

SN - 354039110X

SN - 9783540391104

VL - 4192

T3 - Lecture Notes in Computer Science

SP - 347

EP - 355

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|Lect. Notes Comput. Sci.

PB - Springer Nature

T2 - 13th European PVM/MPI User's Group Meeting

Y2 - 1 July 2006

ER -

Keller R, Bosilca G, Fagg G, Resch M, Dongarra JJ. Implementation and usage of the PERUSE-interface in open MPI. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|Lect. Notes Comput. Sci.. Vol. 4192. Springer Nature. 2006. p. 347-355. (Lecture Notes in Computer Science). doi: 10.1007/11846802_48

Implementation and usage of the PERUSE-interface in open MPI

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this