Customized dynamic load balancing for a network of workstations

Mohammed Javeed Zaki, Wei Li, Srinivasan Parthasarathy

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

42 Citations (Scopus)

Abstract

Load balancing involves assigning to each processor, work proportional to its performance, minimizing the execution time of the program. Although static load balancing can solve many problems (e.g., those caused by processor heterogeneity and non-uniform loops) for most regular applications, the transient external load due to multiple-users on a network of workstations necessitates a dynamic approach to load balancing. In this paper we examine the behavior of global vs local, and centralized vs distributed, load balancing strategies. We show that different schemes are best for different applications under varying program and system parameters. Therefore, customized load balancing schemes become essential for good performance. We present a hybrid compile-time and run-time modeling and decision process which selects (customizes) the best scheme, along with automatic generation of parallel code with calls to a runtime library for load balancing.

Original languageEnglish
Title of host publicationIEEE International Symposium on High Performance Distributed Computing, Proceedings
Pages282-291
Number of pages10
Publication statusPublished - 1996
Externally publishedYes
EventProceedings of the 1996 5th IEEE International Symposium on High Performance Distributed Computing - Syracuse, NY, USA
Duration: 6 Aug 19969 Aug 1996

Other

OtherProceedings of the 1996 5th IEEE International Symposium on High Performance Distributed Computing
CitySyracuse, NY, USA
Period6/8/969/8/96

Fingerprint

Dynamic loads
Resource allocation

ASJC Scopus subject areas

  • Hardware and Architecture

Cite this

Zaki, M. J., Li, W., & Parthasarathy, S. (1996). Customized dynamic load balancing for a network of workstations. In IEEE International Symposium on High Performance Distributed Computing, Proceedings (pp. 282-291)

Customized dynamic load balancing for a network of workstations. / Zaki, Mohammed Javeed; Li, Wei; Parthasarathy, Srinivasan.

IEEE International Symposium on High Performance Distributed Computing, Proceedings. 1996. p. 282-291.

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

Zaki, MJ, Li, W & Parthasarathy, S 1996, Customized dynamic load balancing for a network of workstations. in IEEE International Symposium on High Performance Distributed Computing, Proceedings. pp. 282-291, Proceedings of the 1996 5th IEEE International Symposium on High Performance Distributed Computing, Syracuse, NY, USA, 6/8/96.
Zaki MJ, Li W, Parthasarathy S. Customized dynamic load balancing for a network of workstations. In IEEE International Symposium on High Performance Distributed Computing, Proceedings. 1996. p. 282-291
Zaki, Mohammed Javeed ; Li, Wei ; Parthasarathy, Srinivasan. / Customized dynamic load balancing for a network of workstations. IEEE International Symposium on High Performance Distributed Computing, Proceedings. 1996. pp. 282-291
@inproceedings{2540215b530c40fb85c6792bafcbe25e,
title = "Customized dynamic load balancing for a network of workstations",
abstract = "Load balancing involves assigning to each processor, work proportional to its performance, minimizing the execution time of the program. Although static load balancing can solve many problems (e.g., those caused by processor heterogeneity and non-uniform loops) for most regular applications, the transient external load due to multiple-users on a network of workstations necessitates a dynamic approach to load balancing. In this paper we examine the behavior of global vs local, and centralized vs distributed, load balancing strategies. We show that different schemes are best for different applications under varying program and system parameters. Therefore, customized load balancing schemes become essential for good performance. We present a hybrid compile-time and run-time modeling and decision process which selects (customizes) the best scheme, along with automatic generation of parallel code with calls to a runtime library for load balancing.",
author = "Zaki, {Mohammed Javeed} and Wei Li and Srinivasan Parthasarathy",
year = "1996",
language = "English",
pages = "282--291",
booktitle = "IEEE International Symposium on High Performance Distributed Computing, Proceedings",

}

TY - GEN

T1 - Customized dynamic load balancing for a network of workstations

AU - Zaki, Mohammed Javeed

AU - Li, Wei

AU - Parthasarathy, Srinivasan

PY - 1996

Y1 - 1996

N2 - Load balancing involves assigning to each processor, work proportional to its performance, minimizing the execution time of the program. Although static load balancing can solve many problems (e.g., those caused by processor heterogeneity and non-uniform loops) for most regular applications, the transient external load due to multiple-users on a network of workstations necessitates a dynamic approach to load balancing. In this paper we examine the behavior of global vs local, and centralized vs distributed, load balancing strategies. We show that different schemes are best for different applications under varying program and system parameters. Therefore, customized load balancing schemes become essential for good performance. We present a hybrid compile-time and run-time modeling and decision process which selects (customizes) the best scheme, along with automatic generation of parallel code with calls to a runtime library for load balancing.

AB - Load balancing involves assigning to each processor, work proportional to its performance, minimizing the execution time of the program. Although static load balancing can solve many problems (e.g., those caused by processor heterogeneity and non-uniform loops) for most regular applications, the transient external load due to multiple-users on a network of workstations necessitates a dynamic approach to load balancing. In this paper we examine the behavior of global vs local, and centralized vs distributed, load balancing strategies. We show that different schemes are best for different applications under varying program and system parameters. Therefore, customized load balancing schemes become essential for good performance. We present a hybrid compile-time and run-time modeling and decision process which selects (customizes) the best scheme, along with automatic generation of parallel code with calls to a runtime library for load balancing.

UR - http://www.scopus.com/inward/record.url?scp=0029747931&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0029747931&partnerID=8YFLogxK

M3 - Conference contribution

SP - 282

EP - 291

BT - IEEE International Symposium on High Performance Distributed Computing, Proceedings

ER -