云环境下面向数据密集型应用的容错性资源配置研究  

Fault-tolerant Resource Provisioning for Data-intensive Application in Cloud

作　　者：李宏梅 杨天国 张磊 莫瑞超 许小龙[2,3] 徐占洋 Li Hongmei;Yang Tianguo;Zhang Lei;Mo Ruichao;Xu Xiaolong;Xu Zhanyang(Yunnan Power Grid Corporation Dehong Power Supply Bureau,Dehong 678400,China;School of Computer and Software,Nanjing University of Information Science and Technology,Nanjing 210044,China;Jiangsu Engineering Center of Network Monitoring,Nanjing University of Information Science and Technology,Nanjing 210044,China)

机构地区：[1]云南电网有限责任公司德宏供电局,云南德宏678400 [2]南京信息工程大学计算机软件学院,江苏南京210044 [3]南京信息工程大学江苏省网络监控中心,江苏南京210044

出　　处：《云南电力技术》2022年第1期29-35,39,共8页Yunnan Electric Power

基　　金：国家自然科学基金资助项目(61702277)。

摘　　要：针对云计算平台发生故障所带来的数据密集型应用资源配置问题,为了在云计算平台发生宕机时进行有效的资源配置,本文提出了一种面向数据密集型应用的容错资源配置方法(FRPM)。首先,将数据密集型应用建模为工作流模型,并且基于虚拟层2(VL2)网络拓扑结构建立了宕机任务恢复时间模型和负载均衡模型;然后,利用非支配排序遗传算法(NSGA-III)实现对数据密集型应用的完成时间和云平台所有计算结点负载均衡进行联合优化,从而为部署在宕机结点上应用找到有效的资源配置策略;最后,根据简单加权法(SAW)和多层次决策准则(MCDM)选择出最优的资源配置策略。实验结果表明,FRPM方法能够在云平台在发生宕机时保证数据密集型应用的完成时间最短,与此同时也能保证云平台所有计算结点的负载均衡。The problem of resource provisioning for data-intensive application caused by the failure of the cloud is becoming increasingly prominent.To address the problem of resource provisioning when the computing node of the cloud fails,a fault-tolerant resource provisioning method(FRPM)for data-intensive applications is proposed in this paper.Initially,the data-intensive application is modeled as a workflow model,and the faulty task recovery time model and the load balance model are established based on the Virtual Layer 2(VL2)network topology.Subsequently,the Non-dominated Sorting Genetic Algorithm(NSGA-III)is leveraged to jointly optimize the makespan of data-intensive applications and the load balance of all computing nodes in the cloud to find effective resource provisioning strategies.Finally,the optimal resource provisioning strategy is acquired based on Simple Weighting Method(SAW)and the Multiple Criteria Decision Making(MCDM).The experimental results show that the FRPM method ensure the shortest makespan of data-intensive applications when the cloud platform fails,and at the same time,it can ensure the load balance of all computing nodes of the cloud.

关 键 词：云计算 错误容忍 数据密集型应用 资源配置 非支配排序遗传算法(NSGA-III) 

分 类 号：TM74[电气工程—电力系统及自动化]