The Design And Implementation Of A Multidimensional Report System Based On Mondrian

With the wave of technology development and the advent of big data, more and more companies realize that mining event trends from past business data and as the importance of making decisions based on, but the traditional OLTP(online transaction processing) is not suitable massive data analysis, how quickly the precipitation data from the past to dig out valuable information, and easy to understand way to show to the strategic decision-makers, has become a large technical staff and business owners have been thinking about the problem. Under this background, OLAP(online analysis processing)and data warehouse technology rapid development, has become an effective solution to this problem.Currently, a number of domestic and foreign manufacturers have also launched their own data analysis products, but these tools often enough common, difficult to deploy applications, and expensive, so many companies prohibitive. Based on Mondrian we designed a multidimensional reporting system. The system uses the meta data management and data analysis phase separation design, import and maintenance of metadata management is responsible for the data source, the cube(ie, data cube)customization, and provides cube configuration file format Mondrian Schema data analysis server; data analysis server contains a visual report designer, and dashboard design, allowing users to use drag and drop custom reports and dashboards. In addition,the system also integrates the open source product as a front-end presentation tool Saiku,which support html table or chart format many types of output.This system uses JQuery and Bootstrap front-end framework, based on Spring background with Mondrian development, database using My SQL 5.5, while incorporating the Redis as cache server. The whole system Intelli J IDEA development, based on JDK1.7 version, can be deployed on Tomcat, Jetty, JBoss and other servers.Currently, the system has been successful deployed within a large Internet company,as dozens of products in the company statement analysis services. The company's existing solutions are based on Oracle BIEE, which is costly, and only supports Oracle databases;this system saved the company several hundred thousand dollars of software costs, and provide support for distributed databases such as Hive, supporting data analysis in massive data environment.