Title :
Load Balancing Optimization Using the Algorithm DEPRO in a Distributed Environment
Author :
Nino, Elias D. ; Tamara, Carolina ; Gomez, Karen
Author_Institution :
Dept. of Comput. Sci., Univ. del Norte, Barranquilla, Colombia
Abstract :
In this paper we propose a novel, efficient and practical load balancing algorithm named DEPRO for homogeneous distributed environments. The proposal is presented as a framework based on the layers programming paradigm. The main purpose of the algorithm is the solution of the load balancing distribution problem in systems highly congested by clients. The proposal is compared with the well-known Even Distribution for distributed systems. The comparison is made with performance metrics such as the processing time and the number of lost clients. The results shows that, most of the times, DEPRO presents a better performance than the Even Distribution and in some cases the performance is up to 100%.
Keywords :
distributed algorithms; object-oriented programming; resource allocation; DEPRO algorithm; homogeneous distributed systems; layers programming paradigm; load balancing optimization; performance metrics; processing time; Business; Clustering algorithms; Databases; Delay; Heuristic algorithms; Load management; Programming; delay; distributed systems; layers programming; load balancing;
Conference_Titel :
P2P, Parallel, Grid, Cloud and Internet Computing (3PGCIC), 2012 Seventh International Conference on
Conference_Location :
Victoria, BC
Print_ISBN :
978-1-4673-2991-0
DOI :
10.1109/3PGCIC.2012.39