The improvement and implementation of virtual machine scheduling algorithm in OpenStack

摘要

Cloud computing is a promising technique and OpenStack is the representative open source IaaS platform. The current FCFS model for serving virtual machine requests in OpenStack may suffer the problem of the resource pieces. To address this problem, a two-phase virtual machine scheduling algorithm is proposed in this paper to improve the resource utilization ratio and serve more virtual machine requests. This two-phase scheduling algorithm classifies the status of the system into two different phases and adopts different scheduling strategy respectively. One OpenStack cluster of five compute nodes is built in our lab and a simulation with hundreds of servers is set up. The experiments using this cluster and simulation verify the effectiveness of this algorithm.