System and method for the optimization of database access in data base networks

Abstract

A database network router (DNR) in a database network includes a client access module and a database resource manager that communicate with each other using a common software interface. The client access module communicates with clients including application servers and external database servers; and the database resource manager communicates with the database servers. The DNR also includes a protocol manager for handling multiple database protocols. The database network router also includes local data cache storage to store commonly used data objects, results of parsed query statement, non-optimized query statements, and optimized alternatives and assigns a priority to each client on log-in. The DNR provides a method for connections pooling, a method for balancing the load in a database networks and a method for synchronizing the data stored in cache memory with the database network and method for synchronizing the data stored in cache memory with the database to insure data reliability. The DNR also provides a method for monitoring and optimizing data queries to improve the performance of the database network. The DNR also provides a method to migrate from an old to a new version of the database.

Description

FIELD AND BACKGROUND OF THE INVENTION[0001]The present invention relates to data base networks and an improved method for the access of data and, more particularly, to optimization techniques for the flow of data, to improve the efficiency, reliability and quality of service in database networks.[0002]One or more database servers serve a large number of clients. The term “clients” as used herein includes users, application servers and other databases. A schematic system diagram of a prior art database network is shown in FIG. 1.[0003]In the normal operation of a database network, the clients establish a connection with the database server, and send database requests and database updates to the database server. The database server responds to the requests with data and error messages.[0004]Commonly, there is a high processing load, including read / write load and high memory consumption on a database server, caused, for instance, by massive read requests from its clients. At these time...

AI Technical Summary

Benefits of technology

This patent describes a system and method to make database networks more efficient, reliable, and high-quality. This addresses the problems of current database networks and improves their overall performance.

Problems solved by technology

Commonly, there is a high processing load, including read / write load and high memory consumption on a database server, caused, for instance, by massive read requests from its clients.

At these times, some or all of the clients wait for a long time in response to their data requests, because of limited database resources such as low availability of the database server central processing unit (CPU), low memory availability or a long queue waiting for a response in the database server read / write unit.

Another inefficiency in the operation of current database server networks is related to the management of connections.

Whenever a new client establishes or breaks a connection, a long database process is generated consuming CPU time, memory and read / write resources.

A known cause of inefficiency in database networks is the use of database requests that are not executed using the most optimal execution steps.

Inefficient execution of these database requests causes delay in the response to the client, high CPU consumption, high memory consumption or many read / write requests.

Although algorithms are available that identify problematic database requests and suggest alternatives, these have not been applied in real time applications.

Balance between connections is usually achieved when new connections are established, however, when there are several open connections, some of the database servers may become overloaded.

These overloaded servers have limited performance and long response times.

However, cache servers generally require that the application be aware of them since they do not guarantee synchronization between the cached data and the database.

Images