Check-pointing approach for fault tolerance in OpenSHMEM

Pengfei Hao; Swaroop Pophale; Pavel Shamis; Tony Curtis; Barbara Chapman

doi:10.1007/978-3-319-26428-8_3

Check-pointing approach for fault tolerance in OpenSHMEM

Pengfei Hao
, Swaroop Pophale
, Pavel Shamis
, Tony Curtis
, Barbara Chapman

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

2 Scopus citations

Abstract

Fault tolerance for long running applications is critical to guard against failure of either compute resources or a network. Accomplishing this task in software is non-trivial and there is an added level of complexity for implementing a working model for a one-sided communications library like OpenSHMEM, since there is no matching communication call at the target processing element (PE). In this paper we explore a fault tolerance scheme based on check-point and restart, that caters to the one-sided nature of PGAS programming model while leveraging features very specific to OpenSHMEM. Through a working implementation with the 1-D Jacobi code, we show that the approach is scalable and provides considerable computational resource saving.

Original language	English
Title of host publication	OpenSHMEM and Related Technologies
Subtitle of host publication	Experiences, Implementations, and Technologies - 2nd Workshop, OpenSHMEM 2015, Revised Selected Papers
Editors	Manjunath Gorentla Venkata, Pavel Shamis, Neena Imam, M. Graham Lopez
Publisher	Springer Verlag
Pages	36-52
Number of pages	17
ISBN (Print)	9783319264271
DOIs	https://doi.org/10.1007/978-3-319-26428-8_3
State	Published - 2015
Externally published	Yes
Event	2nd Workshop on OpenSHMEM and Related Technologies, OpenSHMEM 2015 - Annapolis, United States Duration: Aug 4 2015 → Aug 6 2015

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9397
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	2nd Workshop on OpenSHMEM and Related Technologies, OpenSHMEM 2015
Country/Territory	United States
City	Annapolis
Period	08/4/15 → 08/6/15

Funding

This work is supported by the United States Department of Defense and used resources of the Extreme Scale Systems Center located at the Oak Ridge National Laboratory.

Access to Document

10.1007/978-3-319-26428-8_3

Cite this

Hao, P., Pophale, S., Shamis, P., Curtis, T., & Chapman, B. (2015). Check-pointing approach for fault tolerance in OpenSHMEM. In M. G. Venkata, P. Shamis, N. Imam, & M. G. Lopez (Eds.), OpenSHMEM and Related Technologies: Experiences, Implementations, and Technologies - 2nd Workshop, OpenSHMEM 2015, Revised Selected Papers (pp. 36-52). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9397). Springer Verlag. https://doi.org/10.1007/978-3-319-26428-8_3

Hao, Pengfei ; Pophale, Swaroop ; Shamis, Pavel et al. / Check-pointing approach for fault tolerance in OpenSHMEM. OpenSHMEM and Related Technologies: Experiences, Implementations, and Technologies - 2nd Workshop, OpenSHMEM 2015, Revised Selected Papers. editor / Manjunath Gorentla Venkata ; Pavel Shamis ; Neena Imam ; M. Graham Lopez. Springer Verlag, 2015. pp. 36-52 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{d29d28563af348c998c5adcfaa945c28,

title = "Check-pointing approach for fault tolerance in OpenSHMEM",

abstract = "Fault tolerance for long running applications is critical to guard against failure of either compute resources or a network. Accomplishing this task in software is non-trivial and there is an added level of complexity for implementing a working model for a one-sided communications library like OpenSHMEM, since there is no matching communication call at the target processing element (PE). In this paper we explore a fault tolerance scheme based on check-point and restart, that caters to the one-sided nature of PGAS programming model while leveraging features very specific to OpenSHMEM. Through a working implementation with the 1-D Jacobi code, we show that the approach is scalable and provides considerable computational resource saving.",

author = "Pengfei Hao and Swaroop Pophale and Pavel Shamis and Tony Curtis and Barbara Chapman",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2015.; 2nd Workshop on OpenSHMEM and Related Technologies, OpenSHMEM 2015 ; Conference date: 04-08-2015 Through 06-08-2015",

year = "2015",

doi = "10.1007/978-3-319-26428-8\_3",

language = "English",

isbn = "9783319264271",

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

publisher = "Springer Verlag",

pages = "36--52",

editor = "Venkata, \{Manjunath Gorentla\} and Pavel Shamis and Neena Imam and Lopez, \{M. Graham\}",

booktitle = "OpenSHMEM and Related Technologies",

}

Hao, P, Pophale, S, Shamis, P, Curtis, T & Chapman, B 2015, Check-pointing approach for fault tolerance in OpenSHMEM. in MG Venkata, P Shamis, N Imam & MG Lopez (eds), OpenSHMEM and Related Technologies: Experiences, Implementations, and Technologies - 2nd Workshop, OpenSHMEM 2015, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9397, Springer Verlag, pp. 36-52, 2nd Workshop on OpenSHMEM and Related Technologies, OpenSHMEM 2015, Annapolis, United States, 08/4/15. https://doi.org/10.1007/978-3-319-26428-8_3

Check-pointing approach for fault tolerance in OpenSHMEM. / Hao, Pengfei; Pophale, Swaroop; Shamis, Pavel et al.
OpenSHMEM and Related Technologies: Experiences, Implementations, and Technologies - 2nd Workshop, OpenSHMEM 2015, Revised Selected Papers. ed. / Manjunath Gorentla Venkata; Pavel Shamis; Neena Imam; M. Graham Lopez. Springer Verlag, 2015. p. 36-52 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9397).

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

TY - GEN

T1 - Check-pointing approach for fault tolerance in OpenSHMEM

AU - Hao, Pengfei

AU - Pophale, Swaroop

AU - Shamis, Pavel

AU - Curtis, Tony

AU - Chapman, Barbara

N1 - Publisher Copyright: © Springer International Publishing Switzerland 2015.

PY - 2015

Y1 - 2015

N2 - Fault tolerance for long running applications is critical to guard against failure of either compute resources or a network. Accomplishing this task in software is non-trivial and there is an added level of complexity for implementing a working model for a one-sided communications library like OpenSHMEM, since there is no matching communication call at the target processing element (PE). In this paper we explore a fault tolerance scheme based on check-point and restart, that caters to the one-sided nature of PGAS programming model while leveraging features very specific to OpenSHMEM. Through a working implementation with the 1-D Jacobi code, we show that the approach is scalable and provides considerable computational resource saving.

AB - Fault tolerance for long running applications is critical to guard against failure of either compute resources or a network. Accomplishing this task in software is non-trivial and there is an added level of complexity for implementing a working model for a one-sided communications library like OpenSHMEM, since there is no matching communication call at the target processing element (PE). In this paper we explore a fault tolerance scheme based on check-point and restart, that caters to the one-sided nature of PGAS programming model while leveraging features very specific to OpenSHMEM. Through a working implementation with the 1-D Jacobi code, we show that the approach is scalable and provides considerable computational resource saving.

UR - https://www.scopus.com/pages/publications/84952311569

U2 - 10.1007/978-3-319-26428-8_3

DO - 10.1007/978-3-319-26428-8_3

M3 - Conference contribution

AN - SCOPUS:84952311569

SN - 9783319264271

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

SP - 36

EP - 52

BT - OpenSHMEM and Related Technologies

A2 - Venkata, Manjunath Gorentla

A2 - Shamis, Pavel

A2 - Imam, Neena

A2 - Lopez, M. Graham

PB - Springer Verlag

T2 - 2nd Workshop on OpenSHMEM and Related Technologies, OpenSHMEM 2015

Y2 - 4 August 2015 through 6 August 2015

ER -

Hao P, Pophale S, Shamis P, Curtis T, Chapman B. Check-pointing approach for fault tolerance in OpenSHMEM. In Venkata MG, Shamis P, Imam N, Lopez MG, editors, OpenSHMEM and Related Technologies: Experiences, Implementations, and Technologies - 2nd Workshop, OpenSHMEM 2015, Revised Selected Papers. Springer Verlag. 2015. p. 36-52. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-26428-8_3

Check-pointing approach for fault tolerance in OpenSHMEM

Abstract

Publication series

Conference

Funding

Access to Document

Other files and links

Fingerprint

Cite this