摘要
随着过去几十年互联网服务的指数增长,各大网站的访问量急剧上升。海量的用户请求使得热门网站的网络请求率可能在几秒钟内大规模增加。一旦服务器承受不住这样的高并发请求,由此带来的网络拥塞和延迟会极大地影响用户体验。负载均衡是高可用网络基础架构的关键组件,通过在后端引入一个负载均衡器,将工作负载分布到多个服务器来缓解海量并发请求对服务器造成的巨大压力,提高后端服务器和数据库的性能以及可靠性。而Nginx作为一款高性能的HTTP和反向代理服务器,正越来越多地应用到实践中。文中将分析Nginx服务器负载均衡的体系架构,研究默认的加权轮询算法,并提出一种改进后的动态负载均衡算法,实时收集负载信息,重新计算并分配权值。通过实验测试,对比不同算法下的负载均衡性能,改进后的算法能有效提高服务器集群的性能。
With the exponential growth of Internet services in the past few decades,visits to major websites have risen dramatically.Massive user requests make it possible for popular websites to see a massive increase in the rate of web requests in a few seconds.Once the server can’t withstand such high concurrent requests,the resulting network congestion and delays can greatly affect the user experience.Load balancing is a key component of high-availability network infrastructure.By introducing a load balancer in the back end,the workload is distributed to multiple servers to alleviate the huge pressure on servers caused by massive concurrent requests,and improve the performance and reliability of the back-end servers and databases.As a high-performance HTTP and reverse proxy server,Nginx is more and more applied in practice.In this study,we will analyze the architecture of Nginx server load balancing,study the default weighted polling algorithm and propose an improved dynamic load balancing algorithm to collect load information in real time,recalculate and distribute weights.The experiment shows that the proposed algorithm can effectively improve the performance of server cluster by comparing the load balancing performance of different algorithms.
作者
张宇星
马明栋
王得玉
ZHANG Yu-xing;MA Ming-dong;WANG De-yu(School of Telecommunications&Information Engineering,Nanjing University of Posts and Telecommunications,Nanjing 210003,China;School of Geographical and Biological Information,Nanjing University of Posts and Telecommunications,Nanjing 210003,China)
出处
《计算机技术与发展》
2020年第3期73-76,81,共5页
Computer Technology and Development
基金
江苏省自然科学基金-青年基金项目(BK20140868)。
