摘要
针对分布式云端容灾系统中备份服务器的放置问题,提出一种基于整数线性规划(ILP)模型的放置策略。基于服务器共享的思想,在满足传输延迟、链路容量等约束下,以最小化备份服务器数量为目标,构建一个ILP模型并进行求解,获得备份服务器的最优放置方案。对服务器故障进行实时监测,当出现故障时,通过Dijkstra算法获得多跳备用路径,保证容灾系统的可靠性。仿真结果表明,该方案比传统"一对一"备份机制所需的备份服务器数量减少40%以上,有效降低了容灾成本。
For the problem of the placement of backup server in distributed disaster tolerated system,a placement strategy based on integer linear programming（ILP） model was proposed.Inspired by server share ideas,an ILP model was build under consideration of the transmission delay and link capacity constraints to minimize the backup server number.The backup server optimal placement strategy was obtained.The real-time monitoring was made for server fault,and the Dijkstra algorithm was used to obtain multi-hop alternate path when fault occurred,to ensure the reliability of the disaster recovery system.Simulation results show that the scheme can reduce the number of backup servers by at least 40%than the traditional one-to-one backup mechanism,effectively reducing the cost of disaster recovery.
出处
《计算机工程与设计》
北大核心
2017年第7期1848-1853,共6页
Computer Engineering and Design
基金
广东省高等学校优秀青年教师培养计划基金项目(Yq2013171)
关键词
分布式容灾系统
云计算
备份服务器布局
整数线性规划模型
备用路径
distributed disaster tolerant system
cloud computing
backup server placement
integer linear programming
alternate route
