基于软件定义方法的异常流量监测系统

发布时间：2018-06-09 12:18

本文选题：虚拟化 + 云计算　；参考：《南京邮电大学》2017年硕士论文

【摘要】：随着容器技术的兴起,越来越多的企业将docker容器技术应用到生产环境中。为了维持整个环境的安全平稳运行,对运行docker容器的宿主机及其上面运行的docker容器进行监控的需求就日益迫切。本论文首先阐述了当前云计算与虚拟化环境下对于异常流量的监控需求,然后基于openstack环境构建了一套基于软件定义方法的虚拟机监控系统,并实现了当系统中出现异常流量的时候能够实现报警功能。然后基于OpenvSwitch技术构建docker集群的网络架构,基于prometheus+grafana构建开源的docker监控系统。运用Floodlight作为控制器绘制docker网络拓扑并通过sdn方法对ping-flood流量进行了修正,更好的监控运行docker容器的宿主机及其上面运行的docker容器,及时发现一些异常的流量及事件,维护整个生产环境的稳定运行。随着虚拟化技术和docker容器技术的发展,对于虚拟化环境及docker运行环境的监控变得日益迫切。只有对整个系统进行了全方位的监控,才能让我们充分了解整个系统的运行状态,才能更及时的发现问题,维护整个系统安全稳定的运行。
[Abstract]:With the rise of container technology, more and more enterprises apply docker container technology to production environment. In order to maintain the safe and stable operation of the whole environment, it is increasingly urgent to monitor the host running docker container and the docker container running on it. This paper first describes the monitoring requirements for abnormal traffic in the current cloud computing and virtualization environment, and then builds a virtual machine monitoring system based on software definition method based on openstack environment. The alarm function can be realized when abnormal flow occurs in the system. Then the network architecture of docker cluster is constructed based on OpenvSwitch technology, and the open source docker monitoring system is built based on prometheus grafana. Using Floodlight as controller to draw docker network topology and modify ping-flood traffic through sdn method, the host running docker container and docker container running on docker container can be monitored better, and some abnormal traffic and events can be found in time. Maintain stable operation of the whole production environment. With the development of virtualization technology and docker container technology, the monitoring of virtualization environment and docker runtime environment becomes more and more urgent. Only by monitoring the whole system in all directions can we fully understand the running state of the whole system, find problems more timely and maintain the safe and stable operation of the whole system.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP277

【参考文献】