The Technique Research On Performance Optimization In The Centralized Customer Information System Of Commercial Bank

Many domestic commercial banks have built or are building their own customer information systems, which integrate customer information from other systems, achieve single cutomer views and provide realtime and consistent customer information services for other systems. These systems play the core role in the process of customer service, evaluation and analysis within the bank' s IT architecture. They are required to have powerful functions of customer information integrating and efficient facilities of customer information service, as well as stability and reliability for 7*24-hour running.Domestic commercial banks mostly have huge customer groups, especially state-owned commercial banks which have more than one hundred million customers. How to resolve performance problems with the applications on large database? How to establish a system with excellent reliability and self-protecting capability, so that we can ensure the complex IT architecture will not collapse under sudden growth of transaction from internal systems? How to implement the capability of 7*24-hour service and flexible expansion to satisfy the requirements of uninterrupted and long-term accessing to the customer information from all systems of the bank?In this paper, we mainly analyze the questions above and put forward solutions, such as application and data shunting on application-distributed architecture, asynchronous system framework, building flow control mechanism, exception detecting center and logic backup of database, to improve efficiency, stability, expansion ability and resumability of the system.Tests were performed by applying techniques in this paper to a commercial bank CIF system, from which I get the following conclusions: First of all, efficiency of the system is raised remarkably. Secondly, stability and reliability of the system are improved. System keeps running and automatically switches to flow control mode when part of the system fails, while other parts function normally. Thirdly, expansibility of the system is improved. New applications can be easily built in by adding resources or creating new application groups. Lastly, resumability after database collapse is enhanced. It takes only one hour to resume from a collapse of database with 8T of data.