Databases Reference
In-Depth Information
database opened
Finished Duplicate Db at 16-NOV-12
RMAN>
The database duplication is successfully completed. You now have an identical copy of the source database
(
orcl
), named
orcl1
.
How It Works
These prerequisites apply to all database duplication jobs:
nomount
option.
•
You must start the auxiliary instance with the
•
You can't use a standby database as the target database.
•
You'll need the password file for the auxiliary instance only if you're using the RMAN client
on a different host than the auxiliary host or if you duplicate from an active database
(see Recipe 15-4 and Recipe 15-5).
•
The target database can be open or closed.
•
Both the source and destination databases must be on the same operating system platform
(32-bit and 64-bit versions of a platform are considered identical). However, when duplicating
from a 32-bit platform to a 64-bit platform, you must run the
utlrp.sql
script (located in the
$ORACLE_HOME/rdbms/admin directory) to convert the code to the new operating system
format.
If you're using disk-based backups, you can speed up the database duplication by allocating more channels.
If you're using tape backups, however, you can have only as many channels as the number of tape devices you have
available.
You may optionally specify the
password file
option with the
duplicate database
command when duplicating
a database using the backup-based method if you want to make multiple passwords from your target database
available on the new duplicate database. Specifying this option will overwrite any existing auxiliary instance password
file with the source database password file.
■
Note
You don't need a password file when you're duplicating a database on the same host as the source database.
The
duplicate database
command creates the duplicate data files for the new database and recovers them with
the help of incremental backups (if you have any) and archived redo logs. It's your job to ensure that the auxiliary
channel can access all the necessary data file backups and the archived redo logs necessary to create the duplicate
database. You may use a mix of full and incremental backups for this purpose. You can also use the backups of the
archived redo log files or the actual archived redo log files during the database duplication process. By default, RMAN
will duplicate a database from the most recent available backups of the source database and will recover the duplicate
database to the latest SCN contained in the archived redo logs. RMAN can't recover the duplicate database to the
latest SCN, since it doesn't use the online redo logs as part of the database duplication. The RMAN client can run on
any host as long as it can connect to the source and the auxiliary instances. You must ensure that all the necessary
backup files and archived redo log files are accessible by the server session on the duplicate host. The auxiliary
channel performs the bulk of the duplication work through the server session it starts on the host where you are
duplicating the database. The auxiliary channel restores the backups of the primary database and starts the recovery
process as part of the database duplication.
