Optimal deployment and scheduling of elastic cloud services in SaaS scenarios

摘要

The use of virtualization technology makes software applications more scalable and cost effective when they are deployed over cloud computing platforms; but virtualization technology also brings challenges to the resource management and service scheduling. The majority of state of the art research about scheduling of cloud services concentrate on the resource scheduling of IaaS (Infrastructure-asa-Service) cloud. Different with this research, this paper proposes novel mechanisms which provide optimal deployment and scheduling of cloud services in SaaS (Software-as-a-Service) scenarios. We study some differences between IaaS and SaaS scheduling, analyze some limitations of current scheduling mechanisms, and propose a stochastic model to describe the scheduling process as an optimization problem. The most significant property of cloud services - elasticity has been considered in our model, and not only the scheduling of SaaS tasks but also the arrangement of elastic service replicas are provided by our mechanisms. Experiment results show that our deployment and scheduling mechanisms optimize the overall performance of the cloud computing platform.