Performing Basic RMAN TSPITR

Having selected your tablespaces to recover and your target time, you are now ready to perform RMAN TSPITR. You have a few different options available to you:

Fully automated TSPITR--in which you specify an auxiliary destination and let RMAN manage all aspects of the TSPITR. This is the simplest way to perform TSPITR, and is recommended unless you specifically need more control over the location of recovery set files after TSPITR or auxiliary set files during TSPITR, or control over the channel configurations or some other aspect of your auxiliary instance.
Customized TSPITR with an automatic auxiliary instance--in which you base your TSPITR on the behavior of fully automated TSPITR, possibly still using an auxiliary destination, but customize one or more aspects of the behavior, such as the location of auxiliary set or recovery set files, or specifying initialization parameters or channel configurations for the auxiliary instance created and managed by RMAN.
TSPITR with your own auxiliary instance--in which you take responsibility for setting up, starting, stopping and cleaning up the auxiliary instance used in TSPITR, and possibly also manage the TSPITR process using some of the methods available in customized TSPITR with an automatic auxiliary instance.

Fully Automated RMAN TSPITR

When performing fully automated TSPITR, letting RMAN manage the entire process, there are only two requirements beyond the preparations in "Planning and Preparing for TSPITR":

You must specify the auxiliary destination for RMAN to use for the auxiliary set datafiles and other files for the auxiliary instance.
You must configure any channels required for the TSPITR on the target instance. (The auxiliary instance will use the same channel configuration as the target instance when performing the TSPITR.)

RMAN bases as much of the configuration for TSPITR as possible on your target database. During TSPITR, the recovery set datafiles are written in their current locations on the target database. The same channel configurations in effect on the target database are used on the auxiliary instance when restoring files from backup. Auxiliary set datafiles and other auxiliary instance files, however, are stored in the auxiliary destination.

Using an Auxiliary Destination

Oracle recommends that you use an auxiliary destination with your auxiliary instance. Even if you use other methods to rename some or all of the auxiliary set datafiles, specifying an AUXILIARY DESTINATION parameter provides a default location for auxiliary set datafiles for which names are not specified. This way, TSPITR will not fail if you inadvertently do not provide names for all auxiliary set datafiles.

To specify an auxiliary destination, find a location on disk with enough space to hold your auxiliary set datafiles. Then, use the AUXILIARY DESTINATION parameter in your RECOVER TABLESPACE command to specify the auxiliary destination location, as shown in the next section.

Performing Fully Automated RMAN TSPITR

To actually peform automated RMAN TSPITR, start the RMAN client, connecting to the target database and, if applicable, a recovery catalog. This example shows connecting in NOCATALOG mode, using operating system authentication:

% rman TARGET /

Note:

Do not connect to an auxiliary instance when starting the RMAN client for automated TSPITR. If there is no connected auxiliary instance, RMAN constructs the automatic auxiliary instance for you when carrying out the RECOVER TABLESPACE command. (If there is a connected auxiliary instance, RMAN will assume that you are trying to manage your own auxiliary instance, and try to use the connected auxiliary for TSPITR.)

If you have configured channels that RMAN can use to restore from backup on the primary instance, then you are ready to perform TSPITR now, by running the RECOVER TABLESPACE... UNTIL... command.

This example returns the users and tools tablespaces to the end of log sequence number 1300, and stores the auxiliary instance files (including auxiliary set datafiles) in the destination /disk1/auxdest:

RMAN> RECOVER TABLESPACE users, tools 
     UNTIL LOGSEQ 1300 THREAD 1
      AUXILIARY DESTINATION '/disk1/auxdest';

Assuming the TSPITR process completes without error, the tablespaces are taken offline by RMAN, restored from backup and recovered to the desired point in time on the auxiliary instance, and then re-imported to the target database. The tablespaces are left offline at the end of the process. All auxiliary set datafiles and other auxiliary instance files are cleaned up from the auxiliary destination.

Tasks to Perform After Successful TSPITR

If TSPITR completes successfully, you must back up the recovered tablespaces, and then you can bring them online.

Backing Up Recovered Tablespaces After TSPITR

It is very important that you backup recovered tablespaces immediately after TSPITR is completed.

After you perform TSPITR on a tablespace, you cannot use backups of that tablespace from before the TSPITR was completed and the tablespace put back on line. If you start using the recovered tablespaces without taking a backup, you are running your database without a usable backup of those tablespaces. For this example, the users and tools tablespaces must be backed up, as follows:

RMAN> BACKUP TABLESPACE users, tools;

You can then safely bring the tablespaces online, as follows:

RMAN> SQL "ALTER TABLESPACE users, tools ONLINE";

Your recovered tablespaces are now ready for use.

Handling Errors in Automated TSPITR

In the event of an error during automated TSPITR, you should refer to "Troubleshooting RMAN TSPITR". The auxiliary set datafiles and other auxiliary instance files will be left in place in the auxililary destination as an aid to troubleshooting. The state of the recovery set files is determined by the type of failure. Once you resolve the problem, you can try your TSPITR operation again.