arrow
Return

A hierarchical adaptive distributed algorithm for load balancing

delete2004-01-01
delete11
PRE
AI
K
Konstantinos Antonis
J
John Garofalakis
I
Ioannis Mourtos
P
Paul G. Spirakis
DOI:10.1016/j.jpdc.2003.07.002delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
Load balancing/sharing exploits the communication facility between the servers of a distributed system, by using the exchanging of status information and jobs between any two servers of the system in order to improve the overall performance. We propose an adaptive distributed hierarchical scheme, the Virtual Tree Algorithm (VTA), which creates a virtual binary tree structure over the actual network topology. It uses the difference-initiated (DI) technique for load balancing, which needs remote information for the transfer policy and no additional information for the location policy. We demonstrate that the introduced virtual construction keeps the exchanged messages to relatively low levels. To evaluate the performance of our policy, we present both theoretical and simulation results. By using simulation, we compare our results with one of the most representative, adaptive, symmetrical and efficient algorithms, the Variable Threshold (V-THR) algorithm. (C) 2003 Elsevier Inc. All rights reserved.
Keywords:
load balancing
distributed systems
AI Summary

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.

Journal

Journal of Parallel and Distributed Computing cover
Journal of Parallel and Distributed Computing
IF:
4
Papers:
3.8K
Citations:
4.8K

Organization

No organization information available