Large Banks Data Center Host Parallel Batch Research And Practice

The rapid development of banking information has become the financial innovation,institutional innovations as well as a major force in the recycling business. With the growing of the bank's information data's concentration, the resulting risk of financial information process efficiency are becoming increasingly prominent. How to improve the efficiency of batch system and reduce the risk from batch job error, to satisfy the business need of Data Centration, has become a important issue of the Data concentration project for a bank data center.The author has done a lot of research about Tivoli Workload Management, and the efficiency of batch process system. With Analysis of the status and business needs of the data center, the author has put forwarding a feasible method of parallel batch process system, designed the procedure for batch, and participated in the parallel batch project.The author has done a great deal of research about the Parallel Sysplex theory,DB2 Datasharing technology and the Partitioned Tablespace technology. Taking into consideration the characteristics of data center batch system,business need, and the practical feasibility, the author participated in the design of the parallel batch system.Firstly,it greatly improve the efficiceny of batch system by a set of multi-unit batch parallel process based on DB2 Partitioned-tablespace and Partitioned-index instead of one-unit batch process based on DB2 table.Secondly, with the use of Tivoli Workload Management tool, the multiple job process sequence is fixed, which greatly reduce the operation rick,ensure the acurrancy and efficiency of accouting and report.Thirdly, with the implement of error-restart programming and DB2 perfomance adjustment method, it improve the tolerance of batch system and reduce the difficulty of daily maintenance.The parallel batch process system operates very well in Data center. It greatly improve the efficiency and automation of batch system , reduce the risk of accouting error and the time of report transmition from datacenter to branchs of the bank. The successful implementation of the project make it possible to continue the whole data concentration project in the bank.