Research On Cloud Computing Resource Scheduling Algorithm Based On Multistage QoS

At present, Cloud computing is one of the most popular field in computer science. With a few years of development, it has been widely known and used by the market successfully. As a new business model, its great commercial value is also revealing. Cloud computing resource scheduling is the focus of the present study. Extensive researches of resource scheduling algorithm include Min-Min, Max-Min, genetic algorithm, ant colony algorithm, greedy algorithm and simulated annealing algorithm, etc. The guarantee of QoS(Quality of Service) becomes particularly important for cloud computing as a commercial Service providing to users. The QoS evaluation model includes resource performance, reliability, economic applicability and security, etc. The QoS of cloud computing resources involves the basic benefits of the users, at the same time, cloud computing service providers can get the user’s recognition and success by ensuring the QoS. Therefore it’s very important to search the QoS content related of cloud computing resource scheduling.There are some disadvantages in the traditional cloud computing job scheduling strategy. It pays more attention to the efficiency and ignores the quality of service. Based on the characteristics of cloud computing, greedy algorithm and scheduling algorithm in cloud computing, this paper proposes a greedy algorithm based on quicksort model in cloud computing job allocation resource scheduling strategy. This algorithm aimed at several problems of greedy algorithm, which are unordered processing, long waiting time of short job, heavy network and burden load of master node. This algorithm groups the nodes by using the principle of quicksort. Jobs are allocated to the group and distributed to nodes by greedy algorithm. In this way, distribution takes a shorter time with simulating results in close to the greedy algorithm, which guarantee the benefit of QoS of cloud computing service providers and users.Beyond the normal cloud resource usage of the paying customers, cloud service sometimes will produce some surplus resources. To deal with the surplus resources, one way are to transfer the virtual machines together first, and then close the free host in order to reduce energy consumption; another way is to provide free trial for the potential users. Potential users may become paying customers after a period of trial usage, its benefits can be more than energy saving. According to the level of user, the service quality and pattern is different. This paper presents a cloud computing resource scheduling algorithm based on multistage QoS, which provides surplus resources for the potential users at the low rate of paying customers. When the resources of paying users growth, which were transferred from potential users to paying users. It increases the proportion of potential users at the same time of guaranteeing the QoS of paying users.This paper selects CloudSim as the experiment environment and implements greedy algorithm and greedy algorithm based on quicksort model in job allocation which were proposed by this paper. This paper realizes greedy algorithm and greedy algorithm based on quicksort model from the method bindCloudletToVm() in the Class DataCenterBroker. The results show that the algorithm proposed by this paper takes close to greedy algorithm in simulation time and reduces the algorithm running time and improves the QoS guarantee ability of cloud computing service providers effectively. Through the experiment and analysis of cloud computing resource scheduling algorithm of multistage QoS, it is feasible that the cloud computing service providers obtain the potential benefits more than the expenditure of the increasing energy consumption, and it has been proved its effectiveness from the angle of market economy.