基于Docker的动态负载均衡弹性伸缩系统  

Dynamic Load Balancing Elastic Telescopic System Based on Docker

作　　者：王强 王瑞刚[1] 周德永 WANG Qiang;WANG Ruigang;ZHOU Deyong(Xi＇an University of Posts and Telecommunications,Xi＇an 710061;Shaanxi Information Engineering Research Institute,Xi＇an 710061)

机构地区：[1]西安邮电大学,西安710061 [2]陕西省信息化工程研究院,西安710061

出　　处：《计算机与数字工程》2018年第6期1140-1144,1159,共6页Computer & Digital Engineering

摘　　要：通过Docker容器快速实现弹性伸缩,既保证了应用系统的稳定可靠;同时也将Docker容器简单易用,轻快便捷的优势发挥到最大,提高了资源利用率,节约成本。论文采用Telegraf、Influxdb和Grafana组合的监控方案,通过对应用容器的监控,实时掌握应用容器CPU、内存等资源使用情况;Etcd组件作为服务发现机制,提供当前系统可用的容器数量;Confd组件和Haproxy组件做动态负载均衡。实验测试表明,当系统面临大流量、高并发用户请求时,通过本系统弹性伸缩以后,实现负载均衡。为企业构建稳定、高效、可靠的应用系统提供参考。The rapid expansion and retraction of the Docker container ensure that the application system is stable and reliable.At the same time,the Docker container is simple and easy to use,and the advantages of light and convenient use are maximized,re-source utilization is improved,and cost is saved. This article uses the monitoring scheme of the combination of Telegraf,Influxdband Grafana to monitor the application container CPU,memory,and other resource usage conditions in real time through the moni-toring of the application container. Etcd component as a service discovery mechanism to provide the current system the number ofcontainers available. Confd components and Haproxy Components do dynamic load balancing. Experimental tests show that when thesystem is facing large flow and high concurrent user requests,load balancing is achieved after the system is flexibly scaled. Providereference for enterprises to build a stable,efficient and reliable application system.

关 键 词：DOCKER 弹性伸缩 负载均衡 监控 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]