Cloud Computing Environment Based On The Time Margin Of Task Scheduling Algorithm

In the past decades, with the rapid development of computer and Internet technology, information technology has greatly changed the way people live and work. At the same time, users demand computing more and more fast. After distributed computing, parallel computing and grid computing model, the computer industry and academia proposed the cloud computing model. Through the Internet, all kinds of large-scale super cloud computing, storage, integration of resources up to form a virtual pool of computing resources, and rent resources and services to users on demand.Task scheduling is the key technology of cloud computing, with analyzing the local task scheduling, distributed task scheduling and task scheduling for grid, this paper found that cloud computing and grid computing in task scheduling is very similar. So, this paper tries to introduce Max-min algorithm to the cloud environment, and simulate it in Cloudsim. According to the experimental results, Max-min algorithm has the dessert case. On the one hand, when the resources are not enough, the algorithm always priority for the big tasks that make the small tasks can't get good resources or always waiting for the resources. On the other hand, when the resources are enough, the algorithm may occupy too many resources so that every task gets a Vm which not only led to reduce the interest rates of resources, but also led to increase the cost of the cloud platform.The applications of cloud platform are various, and it not only can finish the temporary tasks as grid computing, but also can persistent users deploy applications. How to ensure customer service satisfaction and improve resource utilization has become the focus of cloud computing. Taking the Max-min algorithm could make the resources load balances when the resources are not enough, this paper decided to make an algorithm called Max-min Spare Time (MMST) based on Max-min. MMST algorithm accorded the task completion time constraints on resources, and give priority to the task which have the minimal Spare time. After simulating experiment for MMST algorithm, the paper found that MMST algorithm can finish the tasks on demand and improve the utilization of resources.