Mysql cluster online data redistribution method based on data redundancy and related equipment

Abstract

The invention discloses a mysql cluster online data redistribution method based on data redundancy and related equipment. The method comprises the following steps: receiving a data redistribution request; interrupting a change application of data synchronization between a first main data copy and a first standby data copy in the first node according to the data redistribution request; exporting the physical table space of the first standby data copy to generate a table space file, and respectively copying the table space file to a second main data copy and a second standby data copy of a second node; incrementally synchronizing incremental data stored in the first main data copy after a data synchronization change application is interrupted to the second main data copy and the second standby data copy respectively. According to the method, complex sql splicing conversion does not need to be carried out, the consistency of migrated data is guaranteed, and meanwhile the data redistribution efficiency is improved.

Description

technical field [0001] The invention relates to the field of communication technology, in particular to a data redundancy-based mysql cluster online data redistribution method and related equipment. Background technique [0002] A MySQL database cluster will contain multiple mysql data nodes for data storage. When a new node needs to be added to the cluster, online redistribution of data is required. Common data redistribution algorithms include consistent HASH and virtual mapping tables, etc., to solve how to redistribute data online, that is, which data needs to be moved from which node to which node, but the current data movement usually uses proceed in a logical manner, such as figure 1 As shown, first obtain the data to be moved through SQL query, copy the logical data in full, then insert it into the target node through insert, and finally extract the incremental data and import it to the target node. [0003] Moving data in a logical way mainly has the following dis...

AI Technical Summary

Problems solved by technology

Because the logical process needs to query the data first, this process itself is the process of converting the data stored on the disk into readable text data inside the mysql database, and there is conversion overhead; then splicing insert statements consumes a lot of CPU resources; Finally, inserting into the target stage also requires mysql to convert the readable text data into a disk storage format for storage. This process also requires additional transaction log and binlog log overhead.

[0005] 2. The data text conversion process is complicated and it is difficult to ensure consistency

The process of splicing insert statements needs to consider binary data such as character sets, special characters, translation characters, and blobs; the implementation process is very complicated and prone to omissions

Need to add complex verification logic to check the consistency of data movement, which also further reduces performance

Images