The impact of migration on parallel job scheduling for distributed systems

Yanyong Zhang, Hubertus Franke, Jose E. Moreira, Anand Sivasubramaniam

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

31 Citations (Scopus)

Abstract

This paper evaluates the impact of task migration on gangscheduling of parallel jobs for distributed systems. With migration, it is possible to move tasks of a job from their originally assigned set of nodes to another set of nodes, during execution of the job. This additional flexibility creates more opportunities for filling holes in the scheduling matrix. We conduct a simulation-based study of the effect of migration on average job slowdown and wait times for a large distributed system under a variety of loads.We find that migration can significantly improve these performance metrics over an important range of operating points. We also analyze the effect of the cost of migrating tasks on overall system performance.

Original languageEnglish (US)
Title of host publicationEuro-Par 2000 Parallel Processing - 6th International Euro-Par Conference, Proceedings
EditorsArndt Bode, Thomas Ludwig, Wolfgang Karl, Roland Wismüller
PublisherSpringer Verlag
Pages242-251
Number of pages10
ISBN (Electronic)9783540679561
StatePublished - Jan 1 2000
Event6th International European Conference on Parallel Computing, Euro-Par 2000 - Munich, Germany
Duration: Aug 29 2000Sep 1 2000

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1900
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other6th International European Conference on Parallel Computing, Euro-Par 2000
CountryGermany
CityMunich
Period8/29/009/1/00

Fingerprint

Job Scheduling
Migration
Distributed Systems
Scheduling
Costs
Performance Metrics
Vertex of a graph
System Performance
Flexibility
Evaluate
Range of data
Simulation

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Zhang, Y., Franke, H., Moreira, J. E., & Sivasubramaniam, A. (2000). The impact of migration on parallel job scheduling for distributed systems. In A. Bode, T. Ludwig, W. Karl, & R. Wismüller (Eds.), Euro-Par 2000 Parallel Processing - 6th International Euro-Par Conference, Proceedings (pp. 242-251). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1900). Springer Verlag.
Zhang, Yanyong ; Franke, Hubertus ; Moreira, Jose E. ; Sivasubramaniam, Anand. / The impact of migration on parallel job scheduling for distributed systems. Euro-Par 2000 Parallel Processing - 6th International Euro-Par Conference, Proceedings. editor / Arndt Bode ; Thomas Ludwig ; Wolfgang Karl ; Roland Wismüller. Springer Verlag, 2000. pp. 242-251 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{738ef814c4fa4b8fa4d6012b6af391bc,
title = "The impact of migration on parallel job scheduling for distributed systems",
abstract = "This paper evaluates the impact of task migration on gangscheduling of parallel jobs for distributed systems. With migration, it is possible to move tasks of a job from their originally assigned set of nodes to another set of nodes, during execution of the job. This additional flexibility creates more opportunities for filling holes in the scheduling matrix. We conduct a simulation-based study of the effect of migration on average job slowdown and wait times for a large distributed system under a variety of loads.We find that migration can significantly improve these performance metrics over an important range of operating points. We also analyze the effect of the cost of migrating tasks on overall system performance.",
author = "Yanyong Zhang and Hubertus Franke and Moreira, {Jose E.} and Anand Sivasubramaniam",
year = "2000",
month = "1",
day = "1",
language = "English (US)",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "242--251",
editor = "Arndt Bode and Thomas Ludwig and Wolfgang Karl and Roland Wism{\"u}ller",
booktitle = "Euro-Par 2000 Parallel Processing - 6th International Euro-Par Conference, Proceedings",
address = "Germany",

}

Zhang, Y, Franke, H, Moreira, JE & Sivasubramaniam, A 2000, The impact of migration on parallel job scheduling for distributed systems. in A Bode, T Ludwig, W Karl & R Wismüller (eds), Euro-Par 2000 Parallel Processing - 6th International Euro-Par Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1900, Springer Verlag, pp. 242-251, 6th International European Conference on Parallel Computing, Euro-Par 2000, Munich, Germany, 8/29/00.

The impact of migration on parallel job scheduling for distributed systems. / Zhang, Yanyong; Franke, Hubertus; Moreira, Jose E.; Sivasubramaniam, Anand.

Euro-Par 2000 Parallel Processing - 6th International Euro-Par Conference, Proceedings. ed. / Arndt Bode; Thomas Ludwig; Wolfgang Karl; Roland Wismüller. Springer Verlag, 2000. p. 242-251 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1900).

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

TY - GEN

T1 - The impact of migration on parallel job scheduling for distributed systems

AU - Zhang, Yanyong

AU - Franke, Hubertus

AU - Moreira, Jose E.

AU - Sivasubramaniam, Anand

PY - 2000/1/1

Y1 - 2000/1/1

N2 - This paper evaluates the impact of task migration on gangscheduling of parallel jobs for distributed systems. With migration, it is possible to move tasks of a job from their originally assigned set of nodes to another set of nodes, during execution of the job. This additional flexibility creates more opportunities for filling holes in the scheduling matrix. We conduct a simulation-based study of the effect of migration on average job slowdown and wait times for a large distributed system under a variety of loads.We find that migration can significantly improve these performance metrics over an important range of operating points. We also analyze the effect of the cost of migrating tasks on overall system performance.

AB - This paper evaluates the impact of task migration on gangscheduling of parallel jobs for distributed systems. With migration, it is possible to move tasks of a job from their originally assigned set of nodes to another set of nodes, during execution of the job. This additional flexibility creates more opportunities for filling holes in the scheduling matrix. We conduct a simulation-based study of the effect of migration on average job slowdown and wait times for a large distributed system under a variety of loads.We find that migration can significantly improve these performance metrics over an important range of operating points. We also analyze the effect of the cost of migrating tasks on overall system performance.

UR - http://www.scopus.com/inward/record.url?scp=3342891257&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=3342891257&partnerID=8YFLogxK

M3 - Conference contribution

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

SP - 242

EP - 251

BT - Euro-Par 2000 Parallel Processing - 6th International Euro-Par Conference, Proceedings

A2 - Bode, Arndt

A2 - Ludwig, Thomas

A2 - Karl, Wolfgang

A2 - Wismüller, Roland

PB - Springer Verlag

ER -

Zhang Y, Franke H, Moreira JE, Sivasubramaniam A. The impact of migration on parallel job scheduling for distributed systems. In Bode A, Ludwig T, Karl W, Wismüller R, editors, Euro-Par 2000 Parallel Processing - 6th International Euro-Par Conference, Proceedings. Springer Verlag. 2000. p. 242-251. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).