A Task-oriented Scheduling Algorithm Considering Relative Load and QoS on Private Clouds

摘要

Cloud computing is a computing service mode, and virtualization is the key technology of cloud computing. A good scheduling policy for virtual resources will achieve better utilization of resources and guarantee the QoS of users. A task-oriented scheduling algorithm considering relative load and QoS on clouds (TO-RLQ) is proposed, and the evaluation model of load balancing is put forward in detail. The optimized scheduling algorithm is presented and verified with experiments under the simulated cloud platform-CloudSim. The experiments showed that TO-RLQ would improve the load balancing of the cloud and reduce the violation of QoS with shortening the execution time of the tasks when the number of tasks increased with the high heterogeneous degree.