Adaptive Decentralized Load Sharing Algorithms with Multiple Job Transfers In Distributed Computing Environments

We present a new enhanced version for Adaptive Stable Sender-Symmetric Algorithm which efficiently allocates jobs in the network and possesses multiple job transfer capabilities. Till now, all the algorithms schedule only one job at a time to other nodes in a distributed system. We here present a mechanism that is able to schedule multiple jobs coming to a node to multiple nodes in the network at any instance of the time, with minimum time. Parameters used to find the appropriate node to transfer the job do not solely depend on queue length but also CPU utilization, memory usage and available bandwidth of node. The proposed algorithm works well with different types of topologies. The extensive experimental results demonstrate that our new version of the algorithm outperforms the existing algorithms. The main goal of the algorithm is to minimize flow time and execution time for jobs. The result depends on the number of nodes in the network and the variability of the load