SET

You can configure the basic compression level, which does not require the Advanced Compression Option, by specifying BASIC for the algorithm_name.

If you have enabled the Oracle Database 11g Release 2 Advanced Compression Option, you can choose from the following compression levels:

• HIGH - Best suited for backups over slower networks where the limiting factor is network speed

• MEDIUM - Recommended for most environments. Good combination of compression ratios and speed

• LOW - Least impact on backup throughput and suited for environments where CPU resources are the limiting factor.

Note: The compression ratio generally increases from LOW to HIGH, with a trade-off of potentially consuming more CPU resources.

Since the performance of the various compression levels depends on the nature of the data in the database, network configuration, system resources and the type of computer system and its capabilities, Oracle cannot document universally applicable performance statistics. The decision regarding which level is best must factor in on how balanced your system is regarding bandwidth into the CPU and the actual speed of the CPU. It is highly recommended that you run tests with the different compression levels on the data in your environment. Choosing a compression level based on your environment, network traffic characteristics (workload) and data set is the only way to ensure that the backup set compression level can satisfy your organization's performance requirements and any applicable service level agreements.

Note: V$RMAN_COMPRESSION_ALGORITHM describes supported algorithms.

See Also: Oracle Database Reference entry for $RMAN_COMPRESSION_ALGORITHM.

See Also: Oracle Database Backup and Recovery User's Guide to learn more about this option.

Password-encrypted backups require the correct password to be entered before they can be restored. When RMAN reads an encrypted backup piece, it tries each password in the list until it finds the correct one to decrypt this backup piece. RMAN signals an error if no specified keys work.

Note: If restoring from a group of backups created with different passwords, then specify all of the required passwords on the SET DECRYPTION command. RMAN automatically uses the correct password with each backup set.

See Also: "Encryption of Backup Sets"

The command is useful when stdin and stdout have been redirected. For example, in UNIX you can redirect RMAN's input and output in this manner:

% rman TARGET / < in_file > out_file

By including SET ECHO ON in the in_file, you enable the commands contained in in_file to be visible in out_file.

See Also: "Encryption of Backup Sets"

• Omit IDENTIFIED BY password clause to specify transparent-mode encrypted backups.

• Use IDENTIFIED BY password ONLY to specify password-mode encrypted backups.

• Use IDENTIFIED BY password without ONLY to specify dual-mode encrypted backups.

Create a password that is secure. See Oracle Database Security Guide for more information.

If the password is not surrounded by quotes, then it is translated internally into upper case. Thus, the following clauses are all synonyms for IDENTIFIED BY "SOMEPWD":

• IDENTIFIED BY somepwd

• IDENTIFIED BY Somepwd

• IDENTIFIED BY sOmEpWd

Caution: Wallet-based encryption is more secure than password-based encryption because no passwords are involved. You should use password-based encryption only when absolutely necessary because your backups must be transportable.

See Also: "Encryption of Backup Sets" for details on the different encryption modes

This option overrides settings made with the CONFIGURE ENCRYPTION FOR command. If no data files are configured for encryption, then you must explicitly use ON to encrypt required data files.

If FOR ALL TABLESPACES is not specified, then this setting controls encryption of backups for tablespaces where CONFIGURE ENCRYPTION FOR TABLESPACE tablespace_name has not been used to control encryption behavior.

You can use this command to stage archived redo log files to different locations while restoring a database. RMAN knows where to find the newly restored archived redo log files; it does not require them to be in the destination specified by LOG_ARCHIVE_DEST_1. For example, if you specify a different destination from the one in the parameter file and restore archived log backups, subsequent restore and recovery operations detect this new location.

Use this parameter to restore archived redo log files that are not on disk. RMAN always looks for logs on disk first before restoring them from backups.

RMAN can duplex backups to either disk or tape but cannot duplex backups to tape and disk simultaneously. When backing up to tape, ensure that the number of copies does not exceed the number of available tape devices. Also, if BACKUP COPIES is greater than 1, then the BACKUP_TAPE_IO_SLAVES initialization parameter must be enabled on the target database.

The SET BACKUP COPIES command affects all BACKUP commands in the RUN block issued after SET BACKUP COPIES (but not before) and is in effect until explicitly disabled or changed. The SET BACKUP COPIES command affects only BACKUP commands, but does not apply to the BACKUP AS COPY command.

The SET BACKUP COPIES command affects all channels allocated in the session. The order of precedence is as follows, with settings higher on the list overriding settings lower on the list:

1. BACKUP COPIES

2. SET BACKUP COPIES

3. CONFIGURE ... BACKUP COPIES

The names of the backup pieces are dependent on the FORMAT clause in the BACKUP command. You can specify up to four FORMAT strings. RMAN uses the second, third, and fourth values only when BACKUP COPIES, SET BACKUP COPIES, or CONFIGURE ... BACKUP COPIES is in effect. When choosing which format to use for each backup piece, RMAN uses the first format value for copy 1, the second format value for copy 2, and so on. If the number of format values exceeds the number of copies, then the extra formats are not used. If the number of format values is less than the number of copies, RMAN reuses the format values, starting with the first one.

Note: BACKUP COPIES option is not valid when files are created in fast recovery area. Backups to the fast recovery area cannot be duplexed.

Note: Control file autobackups on disk are a special case and are never duplexed: RMAN always writes one and only copy.

The SET MAXCORRUPT command specifies the total number of physical and logical corruptions permitted in a data file during a backup job. If the sum of physical and logical corruptions detected for a data file is no more than its MAXCORRUPT setting, then the BACKUP command completes. If more than MAXCORRUPT corrupt blocks exist, then RMAN terminates without creating output files.

Whether or not the MAXCORRUPT limit is exceeded, RMAN populates the V$DATABASE_BLOCK_CORRUPTION view with any corrupt block ranges that it finds. However, a backup or restore job is terminated after MAXCORRUPT+1 corrupt blocks are found, so in this case RMAN only records MAXCORRUPT+1 corruptions. Any block corruptions beyond the point at which the backup job terminated are not recorded.

Note: If you specify CHECK LOGICAL, then the MAXCORRUPT limit applies to the sum of logical and physical corruptions detected. Otherwise, MAXCORRUPT only applies to the number of physical block corruptions.

See Also: datafileSpec

This command enables you to change the names for multiple files in the database and provides a quick alternative to naming each file individually (see Example 2-76). This command does not set names for temp files.

When issuing SET NEWNAME FOR DATABASE, you must specify at least one of the following substitution variables to avoid name collisions: %b, %f, and %U. See the semantic entry for TO 'filename' for descriptions of the possible substitution variables.

The new names are used for all subsequent DUPLICATE, RESTORE, SWITCH commands that affect the files in the database and when using the RMAN TSPITR utility. If you do not issue this command before the restore operation, then RMAN restores the files to their default locations.

You do not have to issue the SET NEWNAME commands in any particular order. For example, assume that you run the following series of commands:

SET NEWNAME FOR DATABASE TO '/oradata1/%b';
SET NEWNAME FOR TABLESPACE users TO '/oradata2/%U';
SET NEWNAME FOR DATAFILE 1 TO '+data';

The preceding series of commands is equivalent to the following differently ordered series of commands:

SET NEWNAME FOR DATAFILE 1 TO '+data;
SET NEWNAME FOR DATABASE TO '/oradata1/%b';
SET NEWNAME FOR TABLESPACE users TO '/oradata2/%U';

Note: The SET NEWNAME command supports ASM disk groups.

See Also: formatSpec to learn about substitution variables that are valid in SET NEWNAME FOR DATABASE

This command enables you to specify the names of each data file individually (see Example 2-77). The new names are used for all subsequent DUPLICATE, RESTORE, SWITCH commands that affect the files in the database and when using the RMAN TSPITR utility.

If you run SET NEWNAME FOR DATAFILE and then restore a data file to a new location, then you can run SWITCH to rename the file in the control file to the NEWNAME. If you do not run SWITCH, then RMAN records the restored file as a data file copy in the RMAN repository.

Note: The SET NEWNAME command supports ASM disk groups.

See Also: datafileSpec

This command enables you to change the names for multiple files in the duplicate tablespace and provides an alternative to naming each file of the tablespace individually (see Example 2-77).

When issuing SET NEWNAME FOR TABLESPACE, you must specify at least one of the following substitution variables to avoid name collisions: %b, %f, and %U. See the semantic entry for TO 'filename' for descriptions of the possible substitution variables.

The new names are used for all subsequent DUPLICATE, RESTORE or SWITCH commands that affect the files in the tablespace. If you do not issue this command before the restore operation, then RMAN restores the files to their default locations.

Note: The SET NEWNAME command supports ASM disk groups.

See Also: formatSpec to learn about substitution variables that are valid in SET NEWNAME FOR DATABASE

Note: The SET NEWNAME command supports ASM disk groups.

See Also: tempfileSpec

When issuing SET NEWNAME FOR DATABASE or SET NEWNAME FOR TABLESPACE, you must specify substitution variables within filename to avoid name collisions. Specify at least one of the following substitution variables to avoid name collisions: %b, %f, and %U. The %I and %N variables are optional.

• %b

  Specifies the file name without the fully qualified directory path. For example, the data file name /oradata/prod/financial.dbf is transformed to financial.dbf. This variable enables you to preserve the names of the data files while you move them to different directory. During backup, it can be used for the creation of image copies. The variable cannot be used for OMF data files or backup sets.

• %f

  Specifies the absolute file number of the data file for which the new name is generated. For example, if data file 2 is duplicated, then %f generates the value 2.

• %U

  Specifies a system-generated unique file name. The name is in the following format: data-D-%d_id-%I_TS-%N_FNO-%f. The %d variable specifies the database name. For example, a possible name might be data-D-prod_id-22398754_TS-users_FNO-7.

• %I

  Specifies the DBID.

• %N

  Specifies the tablespace name.

Note: Use of other substitution variables defined in formatSpec is undefined: some yield an error, while others result in zero values and are not supported.

If you set NEWNAME to an ASM disk group for a data file and run RESTORE, then RMAN restores the file to the disk group. If you specify a file name for a temp file, then it is the new name of the temp file after the database is recovered and opened.

For example, SET NEWNAME FOR TABLESPACE users TO NEW sets OMF names for the data files of the users tablespace to be created in the DB_CREATE_FILE_DEST location (see Example 2-78).

If the original file is an Oracle-managed file or is on an ASM disk group, then RMAN attempts to delete the original file. If you specify TO NEW for a temp file, then RMAN creates the temp file in DB_CREATE_FILE_DEST when the database is opened.

See Also: Oracle Database Administrator's Guide for information about Oracle-managed files

Note: You can only use SET TO RESTORE POINT when the database is mounted, because the defined restore points are recorded in the control file. For example, you cannot use SET TO RESTORE POINT to specify the target SCN for a RESTORE CONTROLFILE operation.

See Also: untilClause

Note: The filename is on the host running the RMAN client.

See Also: Oracle Database Reference for more on V$SESSION.CLIENT_INFO

The V$SESSION.CLIENT_INFO column contains information for each RMAN server session. The data appears in one of the following formats:

• id=string

• id=string, ch=channel_id

The first form appears in the RMAN target database connection. The second form appears in all allocated channels. When the current job is complete, the V$SESSION.CLIENT_INFO column is cleared.

See Also: Oracle Database Reference for more on V$SESSION.CLIENT_INFO

1. SET CONTROLFILE AUTOBACKUP executed within a RUN block

2. SET CONTROLFILE AUTOBACKUP executed at the RMAN prompt

3. CONFIGURE CONTROLFILE AUTOBACKUP FORMAT

The %F substitution variable is required to be in the new formatSpec. No other substitution variable is valid in a control file autobackup formatSpec.

See Also: formatSpec for the semantics of the %F substitution variable

RMAN displays the DBID upon connection to the target database. You can obtain the DBID by querying the V$DATABASE view or the RC_DATABASE and RC_DATABASE_INCARNATION recovery catalog views.

You should only run the SET DBID command in the following specialized circumstances:

• You are not connected to a recovery catalog and want to restore the control file (see Example 3-43). The same restriction applies when you use the Data Recovery Advisor to restore a control file autobackup. CONFIGURE can locate an autobackup and restore it only if SET DBID is issued before ADVISE FAILURE.

• You want to restore the server parameter file (see Example 3-44).

• You are connected to the recovery catalog but not the target database and use the FOR DB_UNIQUE_NAME option on the CONFIGURE, LIST, REPORT, SHOW, or UNREGISTER commands.

• You are connected to the recovery catalog and want to restore the controlfile but the target database is not mounted and the database name is not unique in the recovery catalog.

• You are performing a DUPLICATE without a target connection but are not using the DBID subclause in the DUPLICATE command and the database name specified in SET DATABASE or DATABASE clause is not unique in the recovery catalog.

• You have not specified INCARNATION within the DATABASE clause of the DUPLICATE command.

• You want to duplicate to an incarnation not in the current incarnation path (orphan incarnation).