Delay Arrival Scheduling Method Of Cloud Simulation Platform

The rapid growth of Internet users and applications has driven the development of cloud computing.User on-demand access to cloud resources has evolved into a new business model.Various user tasks also bring new challenges to cloud computing.Efficient task scheduling and resource scheduling have become an important research direction of cloud computing.The main task of task scheduling in cloud computing is to calculate and analyze the tasks submitted by users,and assign appropriate virtual resource nodes to them.An excellent scheduling algorithm needs to ensure the stability and robustness of the cloud computing operating environment,as well as shorten the response time to user requests to satisfy customers.This article sets the priority of the task based on the random delayed arrival of the task,and uses the priority as the main scheduling basis.At the same time,the priority of the task will change dynamically over time.In order to prevent the occurrence of starvation of the task,it is necessary to make The overall completion time is relatively small.The first one defines a set from the perspective of task allocation,stores the tasks that have not yet been allocated,and starts to allocate when the set is full,so as to allocate high-speed computing units to tasks with high priority;at the same time,in order to avoid excessive waiting,Define a minimum waiting time to ensure that even if the collection is not satisfied,the task will be allocated within a certain time.The second is to take the task of the virtual machine as the starting point,establish a task queue,determine the dynamic priority based on the initial priority and waiting time of the task and sort them according to the priority level,and each time the task with the highest priority is scheduled for execution.With the increase of time,the initial low-priority task can also become the current higher-priority task,ensuring that the low-priority task can also run.On this basis,the splittable attribute of the task is added to allow idle virtual machines to check whether there is a splittable task being executed,define a simple task splitting strategy,solve the load balancing problem of the virtual machine,and shorten the overall completion time.The research analysis and test results of this paper show that the scheduling method in this paper has a better effect in the system of delayed task arrival,which can not only ensure the early completion of the task,but also deal with the problem of task starvation.