Distributed processing method and system for database

Abstract

The invention discloses a distributed processing method for a database. A service request is processed through the steps that 1, the service request is sent to a proxy server (proxy); 2, the proxy finds out a target server by utilizing a routing algorithm through a locally-cached database server list, and the request is forwarded to an intelligent calculating body (agent) representing the target server; 3, the agent is communicated with the target server through a database adapter, wherein the agent maintains key-value pairs of the target server in distributed task coordination by being regularly communicated with the target server, and the key-value pairs contain an ip address and port of the target server; 4, a processing result is returned to the agent; 5, the agent returns the processing result to the proxy, wherein the proxy maintains a locally-cached database list in real time by monitoring the change conditions of the key-value pairs in distributed task coordination; 6, the proxy returns the processing result to a service. The invention further provides a distributed processing system for the database.

Description

technical field [0001] The invention belongs to the technical field of database application, and in particular relates to a database distributed processing method and a database distributed processing system. Background technique [0002] A database is a warehouse that organizes, stores and manages data according to the data structure. From the way of operating data, databases can be divided into transactional databases and analytical databases. Transactional databases can add, delete, and modify data, and generally serve online businesses; analytical databases are mainly used for analyzing and mining data. , generally used for offline business (data warehouse). [0003] In terms of data structure for storage or processing, databases can also be divided into relational databases, graph databases, key-value databases, document databases, etc. Among them, relational databases can be further divided into row-based storage databases and column-based databases. Storage database...

AI Technical Summary

Problems solved by technology

[0005] 1. Existing distributed solutions are dedicated to a specific database, not portable, or poorly portable

[0006] 2. The existing distributed database data splitting scheme has a great impact on performance

[0007] Due to the problem of data splitting and scattered storage in general distributed databases, the overhead in data positioning, data aggregation, splitting, etc. is relatively high, resulting in the performance of clusters for some small amounts of data, even worse than stand-alone performance.

[0008] 3. Most of the existing distributed database solutions adopt the master-slave mode, which is poor in expansion and maintainability, and easily causes single point of failure

[0009] Existing distributed schemes generally adopt the mode of one master and multiple slaves, which is likely to cause a single point of failure of the master. In addition, in the case of master failure, a new master needs to be elected from multiple slaves, so an election algorithm is also needed. increased system complexity

[0010] 4. In the existing distributed database solutions, most of the business and the database are tightly bound, which is difficult to transplant and has poor flexibility

Images