Research On Cloud Computing Resource Scheduling Strategy Based On Cloudsim

Cloud computing can convert heterogeneous complex resources physically separated from each other to virtual logical unified shared resources, by means of virtualization. distributed computing, resource allocation and management technology.In the resource scheduling model based on load balancing, the process during which the tasks are submitted, obtain the corresponding resources from the data center, and finally are completed, can be subdivided into two levels: task allocation and virtual machine deployment. Task allocation achieves the task to the virtual machine mapping,and virtual machine deployment or migration achieves the mapping from virtual machine to the physical host. In recent years, many domestic and foreign scholars have put forward a lot of resource scheduling algorithms, but each has its advantages and disadvantages. There are still a variety of urgent problems to be solved in this field.In this paper, the task scheduling and virtual machine migration are analyzed respectively for the resource scheduling problem, and the main task allocation and virtual machine migration algorithms are introduced, such as, the random polling algorithm, Min-min algorithm, Max-min algorithm in the aspect of task allocation; pre-copy algorithm and copy algorithm etc. in the field of virtual machine migration.This paper has made two aspects of the work. First, to solve the problem of task assignment, this paper, considering the quality of service and load balancing, puts forward the task allocation algorithm based on optimized ant colony algorithm. The algorithm has modified the visibility of the path nodes, and considers the SLA service assurance and load balancing, to update the pheromone. The optimal solution also ensures that the load average variance is minimum and the completion time is the shortest. Simulation results show that the algorithm can effectively reduce task completion time, and can effectively balance the number of tasks assigned to each virtual machine, which has great practical significance. Second, aiming at the dynamic migration of virtual machine, a virtual machine migration algorithm based on grey prediction model is proposed. In ensuring the quality of service situation, it can predict short-term load changes, dynamically adjust the migration of virtual machine, so as to improve the utilization rate of resources, and can reduce the number of unnecessary migration caused by the instantaneous peak, thus reducing the cost of migration, to provide a reference for the research of the dynamic migration of virtual machines based on prediction method.