基于考试任务的分布式考试系统设计

Design of distributed examination system based on task

为解决传统网上考试系统应用于大规模考试任务时并发访问的瓶颈问题,提高系统的并发访问能力和服务效率,提出了基于考试任务的分布式设计。设计中采用遗传算法,将考试任务自动分解后,根据一定的策略自动分发到各分布式部署的考试服务器中,并行为考生提供考试服务,从而能够充分合理利用软硬件资源,分散系统压力,解决网络带宽、服务器处理能力、数据库访等的瓶颈问题,提高系统并发访问能力。开发了相应的软件系统,并通过实例验证了该设计的有效性,可满足大规模考试任务的要求。

To solve the bottleneck problem that occurred when traditional examination systems used in large-scale test, a distributed testdesign pattem is proposed. By genetic algorithm, the whole test task would automatically be divided into smaller parts and each part would be distributed to the test servers in a distributed deployment, in parallel to provide examination services. It can both effectively solve the network bandwidth, the processor capacity, database access and other bottlenecks, and meet the requirements of distributed real-time and rationality. Finally, the design is validated by practical application.