An integrated approach to parallel scheduling using gang-scheduling, backfilling, and migration

Y. Zhang, H. Franke, J. E. Moreira, Anand Sivasubramaniam

Research output: Chapter in Book/Report/Conference proceedingConference contribution

21 Citations (Scopus)

Abstract

Effective scheduling strategies to improve response times, throughput, and utilization are an important consideration in large supercomputing environments. Such machines have traditionally used space-sharing strategies to accommodate multiple jobs at the same time. This approach, however, can result in low system utilization and large job wait times. This paper discusses three techniques that can be used beyond simple space-sharing to greatly improve the performance figures of large parallel systems. The first technique we analyze is backfilling, the second is gang-scheduling, and the third is migration. The main contribution of this paper is an evaluation of the benefits from combining the above techniques. We demonstrate that, under certain conditions, a strategy that combines backfilling, gang-scheduling, and migration is always better than the individual strategies for all quality of service parameters that we consider.

Original languageEnglish (US)
Title of host publicationJob Scheduling Strategies for Parallel Processing - 7th International Workshop, JSSPP 2001, Revised Papers
EditorsLarry Rudolph, Dror G. Feitelson
PublisherSpringer Verlag
Pages133-158
Number of pages26
ISBN (Print)3540428178
StatePublished - Jan 1 2001
Event7th International Workshop on Job Scheduling Strategies for Parallel Processing, JSSPP 2001 - Cambridge, United States
Duration: Jun 16 2001Jun 16 2001

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2221
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other7th International Workshop on Job Scheduling Strategies for Parallel Processing, JSSPP 2001
CountryUnited States
CityCambridge
Period6/16/016/16/01

Fingerprint

Migration
Scheduling
Sharing
Quality of service
Parallel Systems
Throughput
Response Time
Quality of Service
Figure
Strategy
Evaluation
Demonstrate

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Zhang, Y., Franke, H., Moreira, J. E., & Sivasubramaniam, A. (2001). An integrated approach to parallel scheduling using gang-scheduling, backfilling, and migration. In L. Rudolph, & D. G. Feitelson (Eds.), Job Scheduling Strategies for Parallel Processing - 7th International Workshop, JSSPP 2001, Revised Papers (pp. 133-158). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2221). Springer Verlag.
Zhang, Y. ; Franke, H. ; Moreira, J. E. ; Sivasubramaniam, Anand. / An integrated approach to parallel scheduling using gang-scheduling, backfilling, and migration. Job Scheduling Strategies for Parallel Processing - 7th International Workshop, JSSPP 2001, Revised Papers. editor / Larry Rudolph ; Dror G. Feitelson. Springer Verlag, 2001. pp. 133-158 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{77f8aa9ce4a747b48ef80335d6079cb1,
title = "An integrated approach to parallel scheduling using gang-scheduling, backfilling, and migration",
abstract = "Effective scheduling strategies to improve response times, throughput, and utilization are an important consideration in large supercomputing environments. Such machines have traditionally used space-sharing strategies to accommodate multiple jobs at the same time. This approach, however, can result in low system utilization and large job wait times. This paper discusses three techniques that can be used beyond simple space-sharing to greatly improve the performance figures of large parallel systems. The first technique we analyze is backfilling, the second is gang-scheduling, and the third is migration. The main contribution of this paper is an evaluation of the benefits from combining the above techniques. We demonstrate that, under certain conditions, a strategy that combines backfilling, gang-scheduling, and migration is always better than the individual strategies for all quality of service parameters that we consider.",
author = "Y. Zhang and H. Franke and Moreira, {J. E.} and Anand Sivasubramaniam",
year = "2001",
month = "1",
day = "1",
language = "English (US)",
isbn = "3540428178",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "133--158",
editor = "Larry Rudolph and Feitelson, {Dror G.}",
booktitle = "Job Scheduling Strategies for Parallel Processing - 7th International Workshop, JSSPP 2001, Revised Papers",
address = "Germany",

}

Zhang, Y, Franke, H, Moreira, JE & Sivasubramaniam, A 2001, An integrated approach to parallel scheduling using gang-scheduling, backfilling, and migration. in L Rudolph & DG Feitelson (eds), Job Scheduling Strategies for Parallel Processing - 7th International Workshop, JSSPP 2001, Revised Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 2221, Springer Verlag, pp. 133-158, 7th International Workshop on Job Scheduling Strategies for Parallel Processing, JSSPP 2001, Cambridge, United States, 6/16/01.

An integrated approach to parallel scheduling using gang-scheduling, backfilling, and migration. / Zhang, Y.; Franke, H.; Moreira, J. E.; Sivasubramaniam, Anand.

Job Scheduling Strategies for Parallel Processing - 7th International Workshop, JSSPP 2001, Revised Papers. ed. / Larry Rudolph; Dror G. Feitelson. Springer Verlag, 2001. p. 133-158 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2221).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - An integrated approach to parallel scheduling using gang-scheduling, backfilling, and migration

AU - Zhang, Y.

AU - Franke, H.

AU - Moreira, J. E.

AU - Sivasubramaniam, Anand

PY - 2001/1/1

Y1 - 2001/1/1

N2 - Effective scheduling strategies to improve response times, throughput, and utilization are an important consideration in large supercomputing environments. Such machines have traditionally used space-sharing strategies to accommodate multiple jobs at the same time. This approach, however, can result in low system utilization and large job wait times. This paper discusses three techniques that can be used beyond simple space-sharing to greatly improve the performance figures of large parallel systems. The first technique we analyze is backfilling, the second is gang-scheduling, and the third is migration. The main contribution of this paper is an evaluation of the benefits from combining the above techniques. We demonstrate that, under certain conditions, a strategy that combines backfilling, gang-scheduling, and migration is always better than the individual strategies for all quality of service parameters that we consider.

AB - Effective scheduling strategies to improve response times, throughput, and utilization are an important consideration in large supercomputing environments. Such machines have traditionally used space-sharing strategies to accommodate multiple jobs at the same time. This approach, however, can result in low system utilization and large job wait times. This paper discusses three techniques that can be used beyond simple space-sharing to greatly improve the performance figures of large parallel systems. The first technique we analyze is backfilling, the second is gang-scheduling, and the third is migration. The main contribution of this paper is an evaluation of the benefits from combining the above techniques. We demonstrate that, under certain conditions, a strategy that combines backfilling, gang-scheduling, and migration is always better than the individual strategies for all quality of service parameters that we consider.

UR - http://www.scopus.com/inward/record.url?scp=84944032891&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84944032891&partnerID=8YFLogxK

M3 - Conference contribution

SN - 3540428178

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

SP - 133

EP - 158

BT - Job Scheduling Strategies for Parallel Processing - 7th International Workshop, JSSPP 2001, Revised Papers

A2 - Rudolph, Larry

A2 - Feitelson, Dror G.

PB - Springer Verlag

ER -

Zhang Y, Franke H, Moreira JE, Sivasubramaniam A. An integrated approach to parallel scheduling using gang-scheduling, backfilling, and migration. In Rudolph L, Feitelson DG, editors, Job Scheduling Strategies for Parallel Processing - 7th International Workshop, JSSPP 2001, Revised Papers. Springer Verlag. 2001. p. 133-158. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).