TherChenYang commented on issue #30862: URL: https://github.com/apache/shardingsphere/issues/30862#issuecomment-2060906391
- Read-write separation can only reduce the pressure on the master database. Taking MySQL as an example, when there are a large number of DML and DDL flowing to the slave database, there will be a problem of master-slave delay because the master database can execute concurrently while the Slave SQL Running thread in older versions cannot. If a DDL needs 10 seconds to execute, all other DDL statements have to wait for this one to finish executing before they can continue executing on the slave database. In newer versions of MySQL, the Slave SQL Running thread has been adjusted to allow multiple threads to simultaneously execute SQL from the master database's binlog. - The application scenario of read-write separation is that the reading of data is much larger than the writing of data. If a single-node database is used at this time, the database will be under great pressure at this time. Therefore, the read-write separation strategy is adopted at this time. Since the data The write rate is not very large, using read-write separation at this time can solve the problem well while avoiding master-slave delay -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
