Multidimensional QoS Resource Scheduling Method with Stakeholder Perspective in Clouds

摘要

Cloud computing environment is complex and changeable. It needs to implement the application of large-scale, dynamic, and indefinite cloud users in the form of cloud tasks to a cloud resource environment with physical location separation, heterogeneous resources, and multiple constraints. In order to solve the resource scheduling problem of cloud computing, this paper meets the QoS requirements of cloud users and cloud computing service providers from multiple dimensions, and realizes ultra-large-scale, high-performance and energy-efficient cloud computing resource scheduling. In this paper, a multidimensional QoS cloud computing resource scheduling method with stakeholder perspective of cloud users and cloud computing service providers is implemented, which includes proposing a 2-level cloud computing resource scheduling structure, constructing a multidimensional QoS cloud computing resource scheduling model, designing an MQoS cloud computing resource scheduling optimization algorithm, and simultaneously optimizing multiple objective functions. Performance analysis show that MQoS has obvious advantages in multidimensional QoS performance compared with FIFO algorithm and Genetic algorithm, and achieves good cloud computing system utility to meet the interests of cloud users and cloud computing service providers. MQoS algorithm is also far superior to the traditional algorithms in cloud data center load balancing difference.