NetVault Plug-in for MySQL 11.4 - User Guide

Backing up replication servers

Support for backing up native MySQL Replication environments has the following limitations:

Incremental and Differential Backups on the slave server require that you enable the “--log-slave-updates” option in MySQL. This option tells the slave to log the updates performed by its SQL thread to its own Binary Log. For this option to work, the slave must also be started with the “--log-bin” option to enable Binary Logging. Normally, this option is used to chain replication servers; however, it can also be used for Binary Log backups enabling PIT Recovery of a replicated environment without the complications of purging Binary Logs on the master server before they have been applied to the slaves.

Replication configuration backups

Using the Relay Log Index Path option, you can specify the full path name to the Relay Log Index file to include it in backups. By default, the status files, “master.info” and “relay-log.info,” reside in the same location. If you use the Relay Log Index Path option and the default filenames and locations are retained, the plug-in automatically backs up and restores all these files for a slave replication server.

Restoring replication servers

You can use Full, Incremental, and Differential Backups from the MySQL Replication Slave Instance to perform disaster recovery for the MySQL Replication Master Instance. After the Master Instance has been restored, you can use the same set of backups to restore each Slave Instance to the same level as the Master Instance, and then restart Replication, or you can reinitialize the Slave Instances using other initialization methods provided in the MySQL Reference Guide.

You can use Individual Database/Table backups from both the master and the slave to restore individual databases and tables to the master. If you want to resynchronize an individual table or database on a slave, Quest recommends that you use MySQL’s Replication process for resynchronization instead of restoring to the slave and then trying to get the slave synchronized with the master.

Using the plug-in in a Failover Cluster environment