Rate-Based and Queue-Based Dynamic Load Balancing Algorithms in Distributed Systems

Authors:
Zeng Zeng;Bharadwaj Veeravalli
Affiliations:
The National University of Singapore;The National University of Singapore
Venue:
ICPADS '04 Proceedings of the Parallel and Distributed Systems, Tenth International Conference
Year:
2004

Citing 0
Cited 3

Performance evaluation of network system through UML

ACM SIGSOFT Software Engineering Notes
Minimizing average response time for scheduling stochastic workload in heterogeneous computational grids

HiPC'06 Proceedings of the 13th international conference on High Performance Computing
MLBLM: a multi-level load balancing mechanism in agent-based grid

ICDCN'06 Proceedings of the 8th international conference on Distributed Computing and Networking

Quantified Score

Hi-index	0.01

Visualization

Abstract

In this paper, we classify the dynamic, decentralized loadbalancing algorithms for heterogenous distributed computersystems into three policies: queue adjustment policy(QAP), rate adjustment policy (RAP) and Queue and RateAdjustment Policy (QRAP). We propose two efficient algorithms,referred to as Rate based Load Balancing via VirtualRouting (RLBVR) and Queue based Load Balancing viaVirtual Routing (QLBVR), which belong to the above RAPand QRAP policies, respectively. We also consider algorithmsEstimated Load Information Scheduling Algorithm(ELISA) that was introduced in the literature, to implementQAP policy. Our focus is to analyze and understand the behaviorsof these algorithms in terms of their load balancingability under varying load conditions (light, moderate, orhigh) and the minimization of mean response time of jobs.We compare the above classes of algorithms by a number ofrigorous simulation experiments to elicit their behavior undersome influencing parameters. From these experiments,recommendations are drawn to prescribe the suitability ofthe algorithms under various situations.