Hadoop异构网络下基于自适应蚂蚁算法的策略路由研究

云计算已被我国规划为未来发展的重点项目,该技术将会使我国数千万企业受益。云服务必须高效、快速运行才能发挥其优势。在我国目前有限的带宽基础上,必须研究高效快速的选路机制,并根据各节点最大网络容量来进行资源调度。文章在传统蚂蚁算法上加入了各节点网络容量参数作为阈值进行自适应选路,该算法可使用策略路由的形式在Cisco路由器上应用,不仅能改善Hadoop的资源调度算法在异构环境下效率极低的问题,还够利用蚂蚁算法快速找到最短路径,并能根据路径上节点的网络容量进行调节,从实验情况来看,该方法可有效避免关键路径上的数据拥塞。

Policy Routing Research Based on Adaptive Ant Algorithm on Hadoop Heterogeneous Network

Cloud computing has been regarded as one of the most important planning projects in the future and the technique will be beneficial to thousands of enterprises in our country. The advantages of Cloud service depend on efficient, fast running network conditions. At present, under the condition of limited bandwidth in our country, studying fast and efficient routing mechanism, which Scheduling resource with the maximum capacity of a network node, is necessary. Therefore, the parameters of network capacity is added as the threshold in each node to route adaptively to the traditional ant algorithm. This algorithm can be applied to Cisco routers in the form of policy routing, which improves the extremely low efficiency of resource scheduling algorithm of Hadoop in the heterogeneous environment, finds the shortest path quickly, and can be adjusted according to the network capacity of nodes on the path. The experimental result shows that this method can effectively avoid the congestion of data on the critical path.