On the performance of nearest-neighbors load balancing algorithms in parallel systems

Author

Cortés, A. ; Ripoll, A. ; Senar, M.A. ; Pons, P. ; Luque, E.

Author_Institution

Dept. of Comput. Archit. & Oper. Syst. Group, Univ. Autonoma de Barcelona, Spain

fYear

1999

fDate

3-5 Feb 1999

Firstpage

170

Lastpage

177

Abstract

DASUD (Diffusion Algorithm Searching Unbalanced Domains) is a totally distributed load-balancing algorithm which belongs to the nearest-neighbors class. DASUD detects unbalanced domains (a processor and its immediate neighbors) and corrects this situation by allowing load movements between non-connected processors. DASUD has been evaluated by comparison with two well-known nearest-neighbors load balancing strategies, namely, the GDE (Generalized Dimension Exchange) and the SID (Sender Initiated Diffusion) by considering a large set of initial load distributions. These distributions were applied to ring, tents and hypercube topologies, and the number of processors ranged from 8 to 128. From these experiments we have observed that DASUD outperforms the other strategies used in the comparison as it provides the best trade-off between the balance degree obtained at the final state and the number of iterations required to reach this state

Keywords

parallel processing; performance evaluation; processor scheduling; resource allocation; DASUD; load balancing; load balancing algorithms; nearest-neighbors; parallel systems; Computer architecture; Concurrent computing; Distributed computing; Iterative algorithms; Load management; Operating systems; Parallel processing; Partial differential equations; Plasma simulation; Topology;

fLanguage

English

Publisher

ieee

Conference_Titel

Parallel and Distributed Processing, 1999. PDP '99. Proceedings of the Seventh Euromicro Workshop on

Conference_Location

Funchal

Print_ISBN

0-7695-0059-5

Type

conf

DOI

10.1109/EMPDP.1999.746661

Filename

746661