A malleable-job system for timeshared parallel machines

Laxmikant V. Kale; Sameer Kumar; Jayant Desouza

doi:10.1109/CCGRID.2002.1017131

A malleable-job system for timeshared parallel machines

Laxmikant V. Kale, Sameer Kumar, Jayant Desouza

Computer Science

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

Abstract

Malleable jobs are parallel programs that can change the number of processors on which they are executing at run time in response to an external command. One of the advantages of such jobs is that a job scheduler for malleable jobs can provide improved system utilization and average response time over a scheduler for traditional jobs. In this paper, we present a programming system for creating malleable jobs that is more general than other current malleable systems. In particular, it is not limited to the master-worker paradigm or the Fortran SPMD programming model, but can also support general purpose parallel programs including those written in MPI and Charm++, and has built-in migration and load-balancing, among other features.

Original language	English (US)
Title of host publication	2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002
Publisher	IEEE Computer Society
Pages	230-237
Number of pages	8
ISBN (Print)	0769515827, 9780769515823
DOIs	https://doi.org/10.1109/CCGRID.2002.1017131
State	Published - 2002
Event	2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002 - Berlin, Germany Duration: May 21 2002 → May 24 2002

Publication series

Name	2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002

Other

Other	2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002
Country/Territory	Germany
City	Berlin
Period	5/21/02 → 5/24/02

ASJC Scopus subject areas

Computational Theory and Mathematics
Computer Science Applications
Software
Electrical and Electronic Engineering

Online availability

10.1109/CCGRID.2002.1017131

Library availability

Discover UIUC Full Text

Cite this

Kale, L. V., Kumar, S., & Desouza, J. (2002). A malleable-job system for timeshared parallel machines. In 2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002 (pp. 230-237). Article 1540460 (2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002). IEEE Computer Society. https://doi.org/10.1109/CCGRID.2002.1017131

A malleable-job system for timeshared parallel machines. / Kale, Laxmikant V.; Kumar, Sameer; Desouza, Jayant.
2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002. IEEE Computer Society, 2002. p. 230-237 1540460 (2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002).

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

Kale, LV, Kumar, S & Desouza, J 2002, A malleable-job system for timeshared parallel machines. in 2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002., 1540460, 2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002, IEEE Computer Society, pp. 230-237, 2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002, Berlin, Germany, 5/21/02. https://doi.org/10.1109/CCGRID.2002.1017131

@inproceedings{7f9d474514ca4387bdc4a81e79cf9e77,

title = "A malleable-job system for timeshared parallel machines",

abstract = "Malleable jobs are parallel programs that can change the number of processors on which they are executing at run time in response to an external command. One of the advantages of such jobs is that a job scheduler for malleable jobs can provide improved system utilization and average response time over a scheduler for traditional jobs. In this paper, we present a programming system for creating malleable jobs that is more general than other current malleable systems. In particular, it is not limited to the master-worker paradigm or the Fortran SPMD programming model, but can also support general purpose parallel programs including those written in MPI and Charm++, and has built-in migration and load-balancing, among other features.",

author = "Kale, {Laxmikant V.} and Sameer Kumar and Jayant Desouza",

year = "2002",

doi = "10.1109/CCGRID.2002.1017131",

language = "English (US)",

isbn = "0769515827",

series = "2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002",

publisher = "IEEE Computer Society",

pages = "230--237",

booktitle = "2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002",

note = "2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002 ; Conference date: 21-05-2002 Through 24-05-2002",

}

TY - GEN

T1 - A malleable-job system for timeshared parallel machines

AU - Kale, Laxmikant V.

AU - Kumar, Sameer

AU - Desouza, Jayant

PY - 2002

Y1 - 2002

N2 - Malleable jobs are parallel programs that can change the number of processors on which they are executing at run time in response to an external command. One of the advantages of such jobs is that a job scheduler for malleable jobs can provide improved system utilization and average response time over a scheduler for traditional jobs. In this paper, we present a programming system for creating malleable jobs that is more general than other current malleable systems. In particular, it is not limited to the master-worker paradigm or the Fortran SPMD programming model, but can also support general purpose parallel programs including those written in MPI and Charm++, and has built-in migration and load-balancing, among other features.

AB - Malleable jobs are parallel programs that can change the number of processors on which they are executing at run time in response to an external command. One of the advantages of such jobs is that a job scheduler for malleable jobs can provide improved system utilization and average response time over a scheduler for traditional jobs. In this paper, we present a programming system for creating malleable jobs that is more general than other current malleable systems. In particular, it is not limited to the master-worker paradigm or the Fortran SPMD programming model, but can also support general purpose parallel programs including those written in MPI and Charm++, and has built-in migration and load-balancing, among other features.

UR - http://www.scopus.com/inward/record.url?scp=10044276198&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=10044276198&partnerID=8YFLogxK

U2 - 10.1109/CCGRID.2002.1017131

DO - 10.1109/CCGRID.2002.1017131

M3 - Conference contribution

AN - SCOPUS:10044276198

SN - 0769515827

SN - 9780769515823

T3 - 2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002

SP - 230

EP - 237

BT - 2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002

PB - IEEE Computer Society

T2 - 2nd IEEE/ACM International Symposium on Cluster Computing and the Grid, CCGrid 2002

Y2 - 21 May 2002 through 24 May 2002

ER -

A malleable-job system for timeshared parallel machines

Abstract

Publication series

Other

ASJC Scopus subject areas

Online availability

Library availability

Related links

Fingerprint

Cite this