基于共享内存的路径规划负载均衡算法  

Load Balancing Algorithm for Path Planning Based on Shared Memory

作　　者：邱鲤鲤 姚术林 刘子立 陈云 

机构地区：[1]厦门大学建筑与土木工程学院,福建厦门361005 [2]厦门精图信息技术股份有限公司,福建厦门361008

出　　处：《测绘与空间地理信息》2016年第9期64-67,共4页Geomatics & Spatial Information Technology

摘　　要：为满足前端对路径规划高并发服务需求,本文设计后台采用集群路径规划服务器来实现并行处理,以达到对前置Fast CGI服务器请求进行分流的目的。由此,本文提出基于共享内存的路径规划负载均衡算法,可根据不同路径规划服务器的性能差异,设置相应的加权系数,并按照加权调度算法,将任务交由合适的路径规划服务器来处理,以均衡所有路径规划服务器之间的负载能力,减少服务延迟,从而提高了路径规划服务的稳定性、扩展性、高可用性和执行的效率。To meet the front end on the path planning for high concurrency service needs,this paper designs background using cluster path planning server to realize parallel processing,in order to achieve the diversion for the prepositive Fast CGI server requests.Thus,this paper presents the load balancing algorithm for path planning based on shared memory,which can set the corresponding weighting coefficient based on the differences in the performance of different path planning server,and according to the weighted scheduling algorithm,the task to be processed by the appropriate path planning server.The result shows that this way can not only balance between all path planning server load capabilities,but reduce the server delay,so as to improve the stability of path planning service,expansion,high usability and execution efficiency.

关 键 词：负载均衡 路径规划 共享内存 分布式并行处理 

分 类 号：P208[天文地球—地图制图学与地理信息工程]