服务链联合多元化及动态冗余备份方法研究

Research on joint diversified and dynamic redundancy backup method for service chain

针对电力云环境下虚拟网络函数(virtual network funtion,VNF)单一集偶发系统故障致使服务链(service function chain,SFC)进程开环的难题,在MATLAB仿真实验平台开展了分布式多集和定制冗余解决方法的研究。通过分析异构VNF服务可靠性需求,明确单一集结构微服务部署架构,利用异地部署实现VNF轻型分布,并采用定制冗余策略,进一步提升VNF弹性。结合轻型多集和冗余定制方法的优势,提出了混合整数线性模型的服务链配置新策略。以资源消耗率、服务中断率、服务链可用性为对象参数的实验结果表明,改进算法可有效避免系统故障导致的服务链完全中断,在满足目标服务链可用性水平和提升服务链弹性能力的同时,降低了平台系统多元化及冗余备份成本。

In order to solve the problem of the SFC process opening loop caused by the failure of VNF single cluster system in the power cloud environment,this paper carried out the research on the distributed multi-set and customized redundancy solution on the MATLAB simulation experiment platform.By analyzing the reliability requirements of heterogeneous VNF services,it clarified the single-set deconstruction microservice deployment architecture,used remote deployment to achieve VNF light distribution,and adopted customized redundancy strategies to further improve VNF flexibility.Combining the advantages of lightweight multiple sets and redundant customization methods,this paper proposed a new service chain configuration strategy based on mixed integer linear model.The experimental results with resource consumption rate,service interruption rate,and service chain availability as the object parameters show that the improved algorithm can effectively avoid the complete interruption of the service chain caused by system failures,meet the target service chain availability level and improve the service chain flexibility.At the same time,it reduces the cost of platform system diversification and redundancy backup.