Business processing method, processing node, center node and cluster

Abstract

The invention discloses a business processing method, a processing node, a center node and a cluster. The method comprises the following steps of enabling the processing node of the cluster to receive global snapshot information which is sent by the center node of the cluster at the preset time interval, wherein the global snapshot information comprises a business identification of an business which is being run by the cluster; enabling the processing node to store the received global snapshot information into a cache; enabling the processing node to process the business according to the global snapshot information stored in the cache. The business processing method disclosed by the embodiment has the advantages that the latest global snapshot information is sent to each processing node by the center node of the cluster at the preset time interval, and the global snapshot information can be stored by the processing node; when the global snapshot information is applied in the business processing process, the obtaining of the latest snapshot information from the center node by the complicated information receiving and sending method is not needed, the global snapshot information of the local cache is obtained, the network expense is reduced, and the business processing efficiency is improved.

Description

technical field [0001] The present invention relates to the computer field, in particular to a transaction processing method, a processing node, a central node and a cluster in the computer field. Background technique [0002] Multiple Version Concurrency Control (MVCC) is a mainstream database design mechanism, which is adopted by Oracle database management system, Innode storage engine, SQLServer database management system and PostgreSQL database management system. In order to improve the parallel execution efficiency of the database, the MVCC mechanism will generate a new version when updating records. Therefore, there may be multiple versions of records in the database. Different read and write transactions can see different versions of the records. The read and write transactions do not conflict with each other, and the parallel execution efficiency is high. [0003] A database management system based on the MVCC mechanism needs to obtain a transaction snapshot (Snapsh...

AI Technical Summary

Problems solved by technology

In the case of high load, a large number of transaction snapshot related messages are sent to the central node, and the network is flooded with a large number of messages, which slows down the processing speed of messages. At the same time, the central node needs to process a large number of requests, resulting in slow processing speed, which will greatly Affects the parallel processing performance of the cluster

Images