A Load Distribution through Competition for Workstation Clusters

• Main Authors: SHUM, Kam Hong, Bozyigit, Muslim
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 1994
• Subjects: load distribution; workstation cluster; realistic concurrent application; study deal; dc data; computing system; mapping phase; flexible solution; individual workstation; solution approach; conventional load; independent task; traded commodity; workstation network; elaborate study; subject dc; communication capacity; concurrent application; market data; partitioning phase; Computer Sciences; Numerical Analysis and Scientific Computing
• Online Access: https://ink.library.smu.edu.sg/sis_research/1059; https://ink.library.smu.edu.sg/context/sis_research/article/2058/viewcontent/ShumKH9ISCIS10.1.1.53.6106.pdf
• Institution: Singapore Management University

The aim of this work is to develop a competition driven solution approach for load distribution in distributed computing system (DCS) environments. The subject DCS is composed of a set of workstation clusters. The study deals with concurrent applications, but assumes the existence of the independent tasks executing on individual workstations as well. Akin to conventional load balancing algorithms, the approach considers two phases; the partitioning phase and the mapping phase. Each phase is based on the application and the DCS data which is translated into market data, mainly price. The price is governed by the demand and supply of the traded commodities ( processing and communication capacities) and the competition created between the buyers (tasks) and the sellers (workstations), respectively. The work provides a basis for an elaborate study on realistic concurrent applications and workstation networks. It underlines the development of simple, faster, and flexible solutions to load d...