强负荷网络环境下的大型负载分析平台的设计与优化

发布时间：2019-08-01 07:41

【摘要】：强负荷网络环境下,传统大型负载分析平台存在资源占用率高、效率低等问题。因此,设计了基于Web服务器集群的大型负载分析平台,确保在强负荷网络环境下,通过Web服务器集群提供高可用的服务,实现负载均衡。分析了负载分析平台运行流程,基于该流程设计了负载分析平台的总体结构,平台通过接收模块接收用户申请,再采用分类模块对用户申请进行分类,通过分发模块向用户申请分配相应的进程。信息收集模块将不同服务器节点的运行状态反馈给方案模块,促使方案模块融合神经网络和遗传算法对不同节点的资源分配状态进行设置。分析模块对总体负载分析平台中的不同服务器节点进行监测,分析不同节点是否存在故障以及节点的负载是否均衡。软件设计中给出了负载分析平台模块功能结构,以及信息采集代码设计。实验结果表明,所设计负载分析平台的CPU平均利用率和任务平均处理时间都较优,并且具有较好的吞吐量。
[Abstract]:In the strong load network environment, the traditional large-scale load analysis platform has many problems, such as high resource utilization rate, low efficiency and so on. Therefore, a large load analysis platform based on Web server cluster is designed to ensure that the Web server cluster provides highly available services and realizes load balancing in the strong load network environment. The running flow of the load analysis platform is analyzed. Based on this process, the overall structure of the load analysis platform is designed. The platform receives the user application through the receiving module, then uses the classification module to classify the user application, and distributes the corresponding process to the user through the distribution module. The information collection module feedback the running state of different server nodes to the scheme module, which urges the scheme module to integrate neural network and genetic algorithm to set the resource allocation state of different nodes. The analysis module monitors the different server nodes in the overall load analysis platform, and analyzes whether there are faults in different nodes and whether the load of the nodes is balanced. In the software design, the function structure of the load analysis platform module and the design of the information collection code are given. The experimental results show that the average utilization rate of CPU and the average processing time of tasks are better and the throughput of the designed load analysis platform is better.
【作者单位】：玉林师范学院;
【基金】：国家自然科学基金面上项目(20719662) 广西壮族自治区技术研究项目:网络入侵容忍理论及应用技术研究(自筹项目)(2013LX112) 玉林师范学院校级课题:基于数据挖掘的电商企业客户关系管理应用研究(2014YJYB02)
【分类号】：TP393.0

【相似文献】