Research Of Mass Data Processing In The Telecom Business Analysis Support System (BASS) Based On Cloud Computing Platform

Mobile Business Analysis and Supporting System (BASS in short) is one of core subsystem of China Mobile Business and Operation Supporting System.It has the features of mass data storing,data source complex,processing powerfully and so on. These characteristics make existing mobile BASS face enormous challenges. As a new computing model,Cloud Computing can pool distributed computing resource together to form a huge "resource pool" .It can solve the problem that mass data is difficult to organize and deal with better.In order to apply Cloud Computing in Mobile BASS,many problems must be solve first, how to extract,transform and load (Extract-Transform-Loading ,ETL in short)mass data efficiently is one of these.Accompanied with the features of mobile BASS,this thesis designed a System framework structure based on the thought of China Mobile's Cloud platform―Big Cloud‖,which called Mobile BASS based on Cloud Computing platform (CB-BASS in short).As the reason of the shortcoming of traditional mass data processing algorithms in Mobile BASS, proposed a method called mass data processing based on split mechanism (Split Mechanism Based Data Processing, SMB-DP in short);As the reason of the shortcoming of ETL tasks Scheduling based on greedy algorithm,added the field of priority,proposed a advanced ETL tasks scheduling algorithm called Advanced ETL tasks scheduling based on greedy algorithm (Advanced Greedy Based-ETL , AGB-ETL in short).On this basis, the thesis built the CB-BASS test environment, selected the class of inventory data in the charging and billing test validation in mobile BASS, and then compared and analyzed the test result,in order to verify that SMB-Data Processing algorithm and AGB-ETL Tasks Scheduling algorithm this thesis proposed can do ETL processing with mass data more efficient in Mobile BASS,and satisfied the real acquirement of Mobile BASS more reasonable.