Database transaction processing method and device based on message queue

Abstract

The invention provides a database transaction processing method and device based on a message queue, and the method allows an assigned server to asynchronously consume and process the message; the method comprises the following steps: receiving a database access request; generating an ordered message queue according to the database access request; using a monitor mechanism to execute the message queue processing by sequence until all messages in the message queue are successfully processed or a retry frequency of abnormity messages in the message queue excess a preset threshold; backspacing the executed messages in the message queue when the retry frequency of abnormity messages in the message queue excess the preset threshold.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a method and device for implementing database transaction processing based on message queues. Background technique [0002] In a relational database, when the bottleneck of a single table or single database is reached due to a particularly large amount of data, it is mostly necessary to perform database and table operations on the database. The specific implementation method is to deploy multiple database instances, and at the same time, split the data table horizontally in one database server, thus forming a structure with multiple databases and multiple tables to store multiple types of data. [0003] Generally, the principle of operating a split database or data table is to avoid operating data of several databases at the same time. However, with the gradual increase in business volume, in many cases, some businesses need to operate on multiple databases or data tables at t...

AI Technical Summary

Problems solved by technology

[0006] 1. The existing method requires the transaction coordinator to actively confirm the status of each transaction executor, so it involves multiple network communications between nodes, which takes a long time and consumes a lot of system performance;

[0007] 2. The existing method requires each transaction executor to execute the assigned tasks synchronously, so it is not suitable for a high-concurrency system, and the more database instances, the worse the availability;

[0008] 3. The existing method has high requirements for the fault tolerance of the online environment. Once a problem occurs on a machine, the entire transaction will fail to execute, and partial retrying is not possible.

[0009] To sum up, it is precisely because the existing methods of processing distributed transactions have serious performance problems that most high-concurrency servers are trying to avoid using them, especially those that do not have high time requirements but have high performance requirements. system

Images