The Design And Implementation Of Java-Based Multi-Database Middleware

With the network technology, database technology and enterprise information construction continues to evolve, the coexistence of multiple heterogeneous databases become more common. For example, a company in different departments may use different database systems, which most likely located in different locations and have the characteristics of independence and autonomy. With the increasing demand for enterprise applications, users want to transparently access and process data in multiple databases. However, in the operating environment, data storage, data model and access control strategies differ between heterogeneous databases, resulting in not sharing data and interoperability between the databases,and ultimately making each database into an information island,greatly hindered the flow of data , information sharing in the enterprise, in turn leading to senior corporate decision-making mistakes, causing significant losses of enterprises.Because the system's complexity, the diversity of data structures, enterprises can not abandon all existing heterogeneous databases and rebuild mang new databases.Therefore, without changing the basis of these heterogeneous databases to provide users with a transparent, integrated data access environment, and enable users to a unified model and query processing language to access these databases, become an urgent research topic .To address the intergration problem for multi-database system we propose a solution that is based on the Java multi-database middleware.The program through the use of Java related technologies, XML (Extensible Markup Language) and database related technologies and insert a middleware which is located in front of the client and back-end databases, and then through the middleware to mask the differences between these database systems, and eventually form a single logical database.This will not only maintain the autonomy of these databases, but also allow users like to use as a single relational database to use multiple databases. The main research work and the results are as follows:(1) This paper designs the overall structure of multi-database middleware, which is divided into six functional modules, namely, connection management module, model processing module, query processing module, the transaction processing module, data processing module, and security control module, and the requirements, design and implementation of each of these functional modules are analyzed in detail;(2) By using to achieve a consistent standard JDBC (Java Data Base Connectivity) driver, Non-blocking Java multi-threading, SHA(Secure Hash Algorithm) for asymmetric cryptography, IP (Internet Protocol)technology, and method of virtual database connection pool to solve client and server connection handling, security access and data processing;(3) Shows the method of using javaCC (Java Compiler Compiler) technology to solve Global SQL (Structured Query Language) statement decomposition and uses the pattern mapping tables, rule intermediate tables, and two-phase commit query technology to solve the model conflict, respectively, stored procedures distributed and acrossing multiple databases grouping, aggregation and sorting problems;(4) Through the use of thread monitoring heartbeat technology to achieve dynamic load back-end database, and load balancing;(5) By using two-phase commit protocol and implementing the JTA (Java Transaction Application Programming Interface) specification interface to solve the distributed multi-database middleware transaction processing issues.(6) Finally, the operational status of the middleware is analyzed in detail from various angles, and the middleware is successfully used in a real production environment.(7) By using multi-database middleware, all existing heterogeneous databases and application systems in the enterprises are reused, and information resources are integrated and shared with each other. At the same time, this system design and implementation experience will give other project teams which need to be resolved problem of multi-database integration to play a very good reference.