Design And Implementation Of Distributed Storage Engine In Unified Communication System

Internet OTT business and mobile Internet applications have been rapid popularization,so that people get the way to get information become diversified,the quality of communication services requirements are getting higher and higher,the traditional sense based on voice communication or SMS technical aspects of the communication business has been unable to meet people's daily needs,communications services are also gradually to the voice,video and other multimedia communications direction,and the formation of the traditional communication technology and Internet information technology integration unitified Communication technology,become a hot topic of the field of computer applications.In the aspect of data access of unified communication system,it is necessary to provide a perfect storage engine mechanism for key data such as instant messaging,document and organization structure,while the traditional relational database based on the traditional file system based on the storage of data security,access to data efficiency and follow-up data mining and analysis are not satisfied with the situation,so there is a need to better meet the needs Of the storage service model.With the development of Hadoop technology and Hadoop related subsystems,the advantages of distributed storage are becoming more and more obvious.Based on the analysis of HDFS,MapReduce parallel computing framework and HBase / Hive architecture and their respective characteristics,this paper proposes a new method based on HBase-Hive integrated design of the storage engine design,in order to meet the converged communication system for data security,data acquisition real-time and reliability requirements,and fully study the basic theory of data mining and K-means,PAM clustering algorithm,this paper designs and implements an improved K-means clustering algorithm in combination with MapReduce parallel computing model as a solution to the demand of data mining for converged communication.In the paper structure,this paper first analyzes the background,present situation,significance and related basic technology of the research,combined with the PAM algorithm to improve the K-means algorithm,and then design and implement the distributed storage engine of the various functional modules,finally through the comparative test and performance test and simulation experiments for algorithms,verify the distributed storage engine in the converged communication system in the feasibility and rationality.