Method and device for accessing database, and database system

Abstract

Embodiments of the invention provide a method and a device for accessing a database, and a database system. The database comprises a read-write library and a read-only library. The method comprises: receiving a data access request from a client, the data access request carrying a key value of to-be-accessed data, according to the key value, determining routing list entries corresponding to the to-be-accessed data, wherein the routing list entries include the key value and a target database of the to-be-accessed data, the target database being the read-write library or the read-only library, and sending the data access request to the target database. In the method, through adding a routing function to middleware, the middleware determines the to-be-accessed target database according to the routing list entries, and accurate routing on the to-be-accessed data is performed, and data consistency can be ensured without any modification on the client.

Description

technical field [0001] The embodiment of the present invention relates to database technology, and in particular to a database access method, device and database system. Background technique [0002] In today's increasingly information-expanding world, storage systems are becoming more and more important. Storage systems are mainly divided into relational databases and non-relational databases. In traditional enterprises, relational databases are mainly used, and relational databases are mainly implemented through middleware. Database read-write separation and horizontal expansion (Scalehorizontally, referred to as Scale-Out), but this type of solution can not take into account performance and strong consistency, the application layer itself needs to consider the consistency of application read and write, for traditional enterprise users, it is necessary to modify the application to adapt with this new architecture. [0003] This type of middleware is mainly emerging in Int...

AI Technical Summary

Problems solved by technology

[0002] In today's increasingly information-expanding world, storage systems are becoming more and more important. Storage systems are mainly divided into relational databases and non-relational databases. In traditional enterprises, relational databases are mainly used, and relational databases are mainly implemented through middleware. Database read-write separation and horizontal expansion (Scalehorizontally, referred to as Scale-Out), but this type of solution can not take into account performance and strong consistency, the application layer itself needs to consider the consistency of application read and write, for traditional enterprise users, it is necessary to modify the application to adapt with this new architecture

[0003] This type of middleware is mainly emerging in Internet applications. In order to improve concurrency performance, many Internet scenarios generally choose the final consistent read and write separation, because the use of synchronous replication between writing and reading libraries will restrict the writing performance, so asynchronous replication is generally used , the data written to and read from the database is eventually consistent, and there is a trade-off between performance and consistency. However, when an application requires strong consistency in reading and writing, the application must be required to choose which library to access. Internet application scenarios Most of them do not consider the scenario of achieving strong read-write consistency, and cannot meet the requirements of read-write separation and strong read-write consistency at the same time

Images