Design And Implementation Of Real-timemonitoring Technology In Open-source Cloud Computing Platform

Recently,the rapid development of cloud computing in the information technology field set off a wave of "cloud",which attracts widespread attention at domestic and abroad.Virtualization technology is the core technology of cloud computing,and more and more companies deploy their own cloud data center through virtualization technologies,and provide basic services through virtual machines.This virtualization technology to build cloud data centers significantly improve the utilization of server resources,while also reduce costs.But how efficient and rational use of data center's hardware resources,reduce energy consumption,improve application performance to become a very hard problem.In the process of development of the data center,data center resource,management is seriously flawed,the main problem is the irrational use of resources.Resource utilization is too low,so that the server is empty load,resulting in a lot of waste of resources,increasing data center energy consumption;resource utilization is too high,so that the server is saturated operating state,seriously affecting the user experience,reducing the availability of the data center.Thus,real-time monitoring of data center resources received widespread attention.Through real-time monitoring,understanding of the distribution and utilization of data center resources,so as to solve the data center resource utilization unreasonable and so through a variety of methods.In this paper,we pay attention to the shortcomings of cloud platform,using the popular open source OpenStack cloud platform as the research object.Based on source code analysis,analysis and study of its working mechanism.At the same time,combined with excellent open source virtual machine switches OpenvSwitch and OpenFlow controller NOX,achieving a perfect platform for cloud resources in real-time monitoring component of the cloud platform for real-time resource monitoring and early warning.And on the basis of resource monitoring,to explore solutions to network congestion solutions.A dynamic perception of migration algorithm to achieve the congestion elimination program,through a virtual machine migration,optimize the layout of the network topology,traffic,thereby eliminating network congestion.Finally,OpenStack cloud platform to build a test environment,record and analyze the utilization of network links in the case before and after optimization,resource monitoring program to verify the correctness of the program and eliminate congestion.