An Efficient Load Balancing Algorithm in Distributed Network.

      Аннотация: A distributed system is a communication intensive system; where each system sends and receives some messages to keep update itself with current system status. The communication made to update the system status is an overhead and it can degrade the performance of the system. Reducing the communication overhead in load balancing algorithms to balance a distributed system is a challenging task. In a distributed system the nodes may be of different computing capacity. Assigning the workload to the computing nodes according to their computing capacity will increase the system performance. To assign the workload according to computing capacity there is need to make a priority list of computing nodes. The performance of a distributed system heavily depends on the load balancing algorithm that employed to it. It is beneficial to prioritize the processors based on their computing power and transfer the work load from a heavily loaded system to a lightly laded system. This strategy is able to reduce the response time and provides better resource utilization.

Ключевые слова: Checkpoint, Grid Computing, Job Migration, load balancing, policy, Replication, Scheduling