A database must support data backup and restoration. PolarDB-X provides data restoration at different levels. For example, PolarDB-X supports instance-level consistent backup and restoration, table-level table recycle
bin, and SQL query flashback.
Consistent backup and restoration
The consistent backup and restoration feature allows you to restore the data of an
instance to a specified point in time that is accurate to the second. The consistent
backup and restoration feature of
PolarDB-X is developed based on periodic full backups and global binary logging. For more information
about global binary logging, see
Global data change logging.
For example, a PolarDB-X instance is backed up at 00:00 every Monday, Tuesday, and Friday, as shown in the
preceding figure. If you want to restore the instance data of 14:25:26 on the previous
Sunday, the system selects a full backup set that is generated at a point in time
that is closest to the recovery time objective. The selected point in time is 00:00
of the previous Friday. Then, the system replays global binary logs from 00:00:00
of the previous Friday to 14:25:26 of the previous Sunday to obtain the required data
snapshot.
Table recycle bin
If the table recycle bin feature is enabled, the tables that are dropped or truncated
are temporarily placed in the recycle bin. If a table needs to be restored within
two hours after it is placed in the recycle bin, it can be retrieved from the table
recycle bin in the console.
SQL query flashback
The SQL query flashback feature helps you roll back or partially roll back operations
that are incorrectly performed. For example, if you use a DELETE statement to delete
data by mistake, you can use the SQL query flashback feature to generate a SQL statement
that can recover the data. The SQL statement is an INSERT statement on the deleted
data.