Design And Implementation Of Cloud Service Operation Support System Based On Data Analysis

With the continuous development of cloud computing,more and more Internet enterprises began creating their cloud computing service platform because of its ever-increasing bussiness value.Multi-users oriented public cloud service,as one kind of the cloud service and with the characteristics of providing operation support service by operators,provides service tenants with all kinds of cloud service.For improving the efficiency of the operation work,operators develop operation support system base on their own operation mode and marketing strategy.It seems that the implementation of cloud service operation support system based on data analysis are the trend of the future,for traditional operation mode is unable to provide dynamic adjustment.This paper aims to utilize data analysis technique to support refined and effective operation,which helps to improve the competitiveness of the cloud service product.Firstly the system's business requirements,functional requirements and non-functional requirements are analyzed in detail.Based on the results of requirements analysis,the system that designed and implemented in the paper is specified to consist of data analysis service,tenant management service,product management service and admin permission management service.Then the paper proposes not only the system architecture design scheme based on Browser/Server mode,but also the server-end design scheme based on hierarchy and modularity.Corresponding technical scheme is proposed with Spring MVC,Spr ing,My Batis,My SQL techniques.Data analysis service as one of critial services is implemented with data analysis techniques.The statistic analysis and data visualizatio n techniques are applied to tenants basic portrait analysis,which helps to analyze and show the multi-dimensional information of tenants.And K-means clustering analysis is applied to tenants classification,pointing out the drawbacks of traditional analysis method,the paper proposed an optimization scheme making analysis more accurate.In addition,the implemented system provides tenants arrears alert service,which solves the problem that cloud resource or instances would be recycled if tenants fall into arrears.Morever,this paper proposes a tenants arrear prediciton model,which can be used to predict the malicious arrearage risk of tenants efficiently.At last,the test of the functions and performance of the operation support system was done,it is working properly,suffice the anticipate desire.