Then the control file expands retain the information

SQL keywords and database initialization parameter names. Within script code

examples, I prefer to use lowercase for readability.

}

NOTE The script can be made even simpler by using the BACKUP FULL

specify one, your scripts will work unchanged with Oracle8i and Oracle9i.

To perform the backup, a SYSDBA connection is required. In this case, we connect as SYSDBA using the target�/ command-line option to connect to the Oracle instance

After the backup has completed successfully, you can see information about the backup held in the control file by querying the v$backup_piece view:

select device_type,handle from v$backup_piece;

It’s very helpful to define a naming standard for the backup pieces. The convention used in the example is ORA_ followed by the type of backup (O for open database, C for closed database, and A for archived redo log), followed by the name of the data-base, followed by other information guaranteed to make the name of the piece unique. Structuring the names of your backup pieces might not seem so important now, but it becomes much more significant when you send your backups to a Legato server and you need to distinguish between the different types of backups held on the corporate server, such as NT and UNIX file backups, and Microsoft SQL Server database back-ups. The chosen naming convention immediately identifies the backup piece as belonging to an Oracle database named OMFD1 that is backed up using RMAN.

NOTE RMAN in Oracle9i includes a new command to back up archived redo

run {

allocate channel c1 type disk;