Research On Multi-database Cross-database Operation Technology

Accessing the database requires establishing a database connection.Establishing a connection(which may include some kind of communication connection,starting a service,permission checking,etc.)is itself a time consuming activity,and the system also allocates memory resources.As new applications continue to evolve and the number of users accessing these databases continues to increase,the corresponding access speeds must decline,and even severe server crashes.For each database connection,it must be disconnected after use.Otherwise,if the program fails and closes,it will cause memory leak in the database system and eventually will have to restart the database.Many enterprises have the phenomenon of coexistence of multiple databases.Different database management systems come from different vendors and run on different software and hardware platforms,rslting in many incompatible features between databases,between(heterogeneous)databases.How information is handled effectively is especially important.In practice,if a connection multiplexing technique can be used to achieve multiplexing of database connections,this will improve the operational performance of multiple(heterogeneous)databases.In addition,the use of a common interface to achieve disaster recovery backup,expansion and other management of a variety of databases,is also extremely beneficial technology for ordinary users.This thesis studies a technology that supports the operation of multiple(heterogeneous)database management systems,designs a middleware software model for database common operations,unifies multiple database access interfaces,shields operational differences between databases,and utilizes The multiplexing technology achieves the purpose of connection convergence,completes the cross-database operation of multiple SQL statements on multiple(heterogeneous)databases,and optimizes the performance and security of cross-database operations.This thesis addresses the shortcomings of other data integration methods such as lack of flexibility,high input cost,and long establishment period.By using XML as the public data model of data integration,XML document collection is used as a relay station for data exchange,and middleware is developed to utilize unified The XML view integrates discrete heterogeneous databases to achieve the goal of transparent access to heterogeneous databases and to achieve system portability.In view of the lack of pattern integration concept and the complexity of user operation,the traditional multi-library operation language defines a pattern mapping rule,and solves the pattern conflicts in different pattern integration through field transformation.Finally,a design is based on SQL-92.The kind of SQL language solves the problem that the traditional multi-library language lacks the global mode and the user is overburdened.The cost of query in heterogeneous environment is mainly reflected in the communication cost.This thesis describes the general semi-join optimization algorithm and improves it based on the semi-join algorithm.Through analysis and experiment,the improved optimization algorithm is compared with the general one.The semi-join algorithm improves the data transmission amount and query response time significantly,and improves the query access efficiency.