Research On Task Schedule In The Platform Of Cloud Testing Platform

Test is the one of the most important procedures among the software development process. At present, there are many methods about improving the test efficiency and the test quality and the Cloud Testing is one of them which are based on Cloud Computing. Cloud Testing has many advantages, such as accessibility, convenience, low cost. But there are many disadvantages, such as big energy waste, simplicity of task schedule.The energy conservation not only protects our environment, but also saves the cost of enterprise. For the purpose of energy conservation, the paper studies the virtual machine schedule in the platform of cloud testing. It is feasible method to improve the resource utilization and shut down the redundant physical machine. So the paper designs the five statuses of physical machine, proposes the conception of modified resource utilization, designs the method of computing the modified resource utilization of CPU, network and physical machine. The paper proposes one computing formula which can be used to calculate the cost of energy and links the modified resource utilization and cost of energy. Finally, the paper proposes the test virtual machine migration schedule algorithm. It is shortened to TVMSA.With various tasks running on the platform of test cloud, improving the efficacy of task schedule not only decrease the cost time, but also increase the users' satisfaction. For the purpose of decrease the cost time, the paper studies task schedule. The paper designs one new kind of method of population initialization, designs the coding map scheme and the possibility of crossover and variation, and provides their computing formula. Finally, the paper proposes the task schedule algorithm that aims at improving efficacy of task schedule.CloudSim, which is simulation software about computing cloud, is used to verify algorithms that are proposed in the paper. By comparing TVMSA and FFD algorithm, TVMSA has superiority in the number of running physical machine, cost energy and average CPU utilization. By comparing task schedule algorithm, RR and GA, the task schedule algorithm proposed in paper is strong in the diversity of population, the stability of population, and cost execution time of tasks.