Design Andimplementation Of Cloud Network Connectivity Quality Analysis System

As the scale of the Internet continues to expand,the traditional network architecture can no longer meet the needs of diversified network applications.In order to carry more business needs,SD-WAN(Software Defined Wide Area Network)technology breaks through the traditional Network mode and becomes a new public cloud networks solution.However,how to provide users with high-quality and stable network services is a problem that public cloud providers are facing,and the operation and maintenance of public cloud networks will face enormous challenges.In this thesis a network data visualization analysis system for public cloud network operation and maintenance personnel was designed and developed.The measurement data of the network layer from the perspective of the application layer was analyzed.Based on the application layer’s demand for network service quality,in the face of millions of measurement data at the network layer,the efficient processing and analysis solutions were designed,and the visual presentation of data was designed.By using this system,users can analyze cloud network status,network service quality,and connection quality,etc.,for public cloud network operations.Maintenance personnel provide decision support in the operation and maintenance work.Firstly,in this thesis the functional and non-functional requirements of the cloud-network connectivity quality analysis system was analyzed.Secondly,the overall architecture design of the system was carried out,and the modular design and development are carried out based on the Spring Cloud microservice technology to achieve system decoupling.Thirdly,carried out the detailed design and implementation of the system functional modules,and show some front-end page implementation renderings,the system test plan was designed and the test results was analyzed,and finally,summarize and look forward to the work of the thesis.The front-end of the system used the Vue.js framework and BootstrapVue supported by Vue as the UI framework,the back-end was developed in Java,and the Redis cache technology was introduced as an in-memory database to store the processed network measurement data.The main contents of this article include:(1)The processing method of dynamically increasing millions of magnitude and disordered public cloud network measurement data was designed,based on the requirements of network data analysis,efficient storage and read schemes were designed.(2)Based on the user’s analysis requirements for network data,in this thesis two virtual private line calculation algorithms were designed,one was used to calculate the virtual private line between the designated data centers that meets the specific network service quality requirements of the application layer,and the ideas to improve computing efficiency in massive data scenarios were designed;Another algorithm was used to calculate the optimal virtual private line between any two data centers in the FULL MESH topology.This algorithm was based on the traditional two-stage algorithm and combined with the calculation scenario of this thesis,and proposed to determine the node pair calculation based on the statistical historical calculation result information.Priority,thereby reducing the number of calculation iterations in the two-stage algorithm and reducing the calculation time of the algorithm;(3)Based on the requirements of public cloud network operation and maintenance personnel for network data visualization,the implementation architecture of the whole system was designed,and a high stability,high availability of cloud network connectivity quality analysis system was developed.