Profit-driven Cloud Service Request Scheduling Under SLA Constraints

摘要

As cloud computing becomes widely deployed, more and more cloud services are offered to end users in a pay-as-you-go manner. How to schedule the dynamic user service requests more cost-effectively with less SLA violations is one of the most intractable problems of cloud service providers. To deal with this challenge, we first establish a cloud service request model with SLA constraints, and then present a new optimization algorithm for profit-driven service request scheduling based on dynamic reuse, which takes account of the personalized SLA characteristics of user requests and current system workload. Our proposed algorithm builds a dynamic virtual machine resource pool on demand, achieves optimal cloud service request scheduling in reasonable time, and thus significantly reduces operational costs of cloud service providers and increase profits of cloud service providers. Our simulation experiments show that our proposed algorithm improves virtual resource utilization and increases profits of cloud service providers compared with several baseline algorithms.