Modelling, evaluating and designing virtual machine scheduling by a clustering mechanism in cloud computing environments

Abstract

Cloud computing is still in its infancy in spite of recently gained significant momentum. To allocate virtual machine resources in cloud data centre and taking both cloud preference and fairness into account are some of the major challenges. In this paper, a novel virtual machine scheduling algorithm by a clustering mechanism for maximising cloud computing system utility is put forward with a brief survey of virtual machine scheduling suitable for large-scale distributed virtual computing environments. It includes (a) modelling a general virtual machine scheduling model for clouds; (b) evaluating cloud preference of user and realising pre-classification virtual machines resources by a clustering mechanism in cloud data centre. The clustering mechanism is applied to meeting performance and fairness requirements of users and (c) designing the virtual machine scheduling algorithm in a cloud. Theoretical as well as experimental results conclusively demonstrate that the scheduling algorithm has high potential as it takes both preference and fairness into account, and maximises cloud computing system utility by the clustering mechanism in cloud computing environments.