Design And Implementation Of Telecom Operation And Maintenance Capacity Analysis System

With the rapidly development of computer science and information technology,enormous data is constantly generated.For enterprises in urgent need of changing their traditional architecture,how to analyze these data and effectively use the value of data is a very important problem for enterprises in the face of massive data.At the same time,how to optimize enterprise service has become an inevitable problem in procedures of modern enterprise transformation.Therefore,in view of the above problems,people propose a machine learning based brain power manipulation and maintenance(aiops)technology to solve these difficulties.Capacity analysis is an extremely important part of intelligent operation and maintenance(aiops).The telecom operation and maintenance capacity analysis system studied in this thesis can majorization and adjust the frame in the company's system resources,solve the problem of system resource allocation to a certain extent,and greatly save time and labor costs.With the help of artificial intelligence technology,the capacity analysis system uses the existing machine learning model to make an assay of and model the numerous digital information collected,so as to solve some problems that can not be solved before.Based on the above background,a capacity analysis system for telecom operation and maintenance in Hadoop environment is designed and implement.Combined with the latest research content at home and abroad,the main research contents of this thesis are as follows:(1)A telecom operation and maintenance capacity analysis system is designed and implement based on Hadoop.Firstly,the requirement of the capacity analysis system is designed.In terms of the requirement of the system,the overall infrastructure,technical infrastructure and system function flow plan are worked out.Finally,based on the SpringBoot framework,each functional block of the system are achieved in particular,and the implementation consequences and some significant code screenshots are given.(2)Hadoop framework and HDFS are used to solve the problem of massive log data storage.Using message middleware and distributed data collection system to memory a large number of data,the matter on real-time memorize of quantities information is solved.Build Hive data warehouse to solve the problem of log data maintenance.(3)SpringBoot is used to integrate MapReduce computing engine,and MapReduce program is written in Java language to analyze and calculate massive log data,which solves the problems of memory overflow and data skew in the process of MapReduce computing(4)Combined with some components of Spring Cloud,the system module is decoupled and the service is called remotely,which increases the availability of the system.