Abstract

With the increasing growth of computing technology, the computers have become more efficient and compact. With the development of powerful microcomputers, the distributed computing has been introduced as an optimal solution to gain faster computing using existing capacities. One of the most important issues in distributed computing systems is allocating a set of tasks to a set of processors to create the load balance and to minimize the total costs. Due to the large scale and complexities of distributed systems, centralized allocation of tasks to specific servers is impossible. To properly manage the service providing resources, we need to acquire the load balance to be proposed to the service provider. Components are regularly monitored and when one component is not responding, the load balancer pops-up and prevents the traffic from being driven to that component. With proper in-situ load level analysis, problems of resource consumption can be usually be minimized, which not only leads to lower costs and green computation, but also lowers the pressure on particular circuits and potentially lengthens their service life. In fact, the aim of load balancing is to find a suitable mapping of the tasks on the processors in the system‎.