Design And Implementation Of Task Scheduling Algorithm Based On Quality Of Service

The rise of cloud computing has injected vitality into the software and hardware, communication and Internet of things. Driven by the business and academia, cloud computing is developing rapidly and has been applied to governments, schools, enterprises and so on. The traditional computing model and service pattern has changed greatly with the appearance of cloud computing and a new business model has been formed. Due to the business characteristics of cloud computing, it should focus on user requirements and improve its service quality. The online service mode of cloud computing should take some security measures to safeguard the interests of the service providers and customers, so the evaluation of quality of service is particularly important. Task scheduling is one of the key tasks of cloud computing, which is to choose the appropriate resources for the job that users submit to be carried out. The system performance and the quality of service are the two factors that should be considered. Research work based on the two factors is done as follows:(1) A cloud computing task scheduling algorithm T-C-TSA (Time-Cost-Task Scheduling Algorithm) has been introduced considering the quality of service of cloud computing. Taking into account the completion time and cost of user tasks, the algorithm extends the task model by increasing the priority factor of user tasks. Meanwhile the virtual machine resources model has been extended. Each task has been scheduled to the most suitable virtual machine by using the thought of traditional greedy scheduling algorithm. The task scheduling algorithm is tested in CloudSim simulation environment and the results prove that T-C-TSA realize the goal of scheduling, ensuring the task scheduling span and less expense.(2) The T-C-TSA algorithm is scheduling task to the virtual machine which owes minimum compromise value of completion time and cost, this method could overload the virtual machine. The LB-TC-TSA(Load balancing-Time-Cost-Task Scheduling Algorithm), which is an improved T-C-TSA algorithm, is come up with to solve this problem. The target of load balancing is added to the improved algorithm, not only considering the task scheduling time-cost compromise values, but considering the load of virtual machine.The improved scheduling algorithm is tested using the simulation tool CloudSim, the results show that the algorithm can achieve good load balancing effect while ensuring the quality of service of cloud computing.