Research on a Performance Enhancement Method for Heavy Load Network

摘要

Response efficiency, load-carrying capacity, and intelligent fault-tolerant capability are three main performance bottlenecks which restrict the overall performance of the information network, so that the application system in the network is difficult to achieve the desired satisfactory results. Aiming at the above problems, this paper proposes a new network structure based on multi-node continuous collaboration and intelligent fault tolerance. In the new structure network, the user is divided into query group and read-write group according to the difference of user's permission to operate database. The user's access load in each group is handled by a separate task processing system. The task processing system consists of gateway group, communication link, application server group, database server group, and database. The real-time data synchronization problem is solved by data synchronization mechanism between databases. The problem of fault redundancy and load balancing is solved by cluster technology. The test results show that the information network performance bottleneck is greatly eliminated, the response efficiency of heavy load interval is more than 50 higher than that of traditional structural network, and the carrying capacity and intelligent fault tolerance capacity are obviously improved.