Balanced workload clusters for distributed object oriented software

Abstract EN

When clustering objects to be allocated on a number of nodes, most researches focus only on either the communication cost between clusters or the balancing of the workload on the nodes.

Load balancing is a technique to distribute workload evenly across two or more computers, network links, CPUs, hard drives, or other resources, in order to get optimal resource utilization, maximize throughput, minimize response time, and avoid overload.

In this paper, we introduce three clustering algorithms that obtain balanced clusters for homogeneous clustered with minimized communication cost.