MySQL Backup provides capabilities for backing up and restoring databases. MySQL Backup includes these features:

A backup operation can include tables for different storage engines and the backup image will still be consistent. That is, you need not care which storage engines you're using. BACKUP DATABASE saves the data in a consistent backup image with respect to its “validity point.”

The validity point ties the backup to the binary log. Restoring a backup can be combined with use of the binary log to accomplish point-in-time recovery: If the restore operation is done because data loss has occurred after the backup was made (that is, after the validity point), restored databases can be brought up to the time of data loss by executing the data changes in the binary log between the times when the backup was made and when the data loss occurred.

A goal of the BACKUP DATABASE and RESTORE statements is to enable other database operations to proceed concurrently, to make it unnecessary to take databases offline or prevent clients from accessing them. BACKUP DATABASE must block some operations from occurring (such as dropping tables from a database while it is being backed up), but the attempt is made to keep blocking to a minimum. Generally, blocked operations are those involving Data Definition Language (DDL) statements. RESTORE must do more blocking because it writes database contents rather than just reading them.

The following discussion covers these aspects of MySQL Backup:

Other parts of this manual provide additional information about MySQL Backup: