Creating an Oracle Clone Environment Using Application-Aware Backup Snapshots
You can duplicate the source database by cloning the snapshots that were created during the backup. You can clone a database to the same server where the source database resides or to a different server.
Cloning is only supported for snapshots of VMware Application-Aware backups on Linux VMs.
The cloning feature uses the storage array hardware capability. For information on the supported arrays, see Instant Clone for Applications.
You can use the cloning feature to:
- Duplicate large databases in a short amount of time
- Create test environments
- Retrieve data quickly without intensive resource use on the production environment
- Alleviate the load on production servers
You can create a clone from an existing data and log volume full backup. When you create a clone, you must specify the reservation period. The system automatically shuts down the clone database and frees up all resources, including the snapshot clones on the storage array, when the reservation period expires.
Schedule clones to run periodically so that the clones are refreshed on a regular basis.
For information on how to clone Oracle 12c PDBs, see Cloning an Oracle 12c Pluggable Database (PDB).
Note: You can create clones of only non ASM databases from Appaware backup snapshots.
Before You Begin
- Verify that the Oracle database version on both the source database server and the destination database server are the same.
- Verify that the Oracle user in both the source database server and the destination database server have the same user ID.
- Verify that the Oracle Agent and the MediaAgent packages are installed on both the source and destination database servers.
- On the destination database server where you want to create the clone database, create a directory for the mount path for the database clone files. This directory must be writable by the Oracle user.
- Do one of the following:
- If the Oracle database version is 18.104.22.168.0, 22.214.171.124.1, or 126.96.36.199.2, install Oracle patch 13366202 or the latest Oracle 11gR2 Patch Set Update on the destination database.
- If the Oracle database version is Oracle 12c, install at a minimum Oracle 188.8.131.52 with patch 19404068.
- The clone operation creates the control file and temp files in the staging path, by default. If you want to change the location of the temp files, create a custom pfile and add the new path to the db_create_file_dest parameter in the pfile.
- Optional: If you want to create a clone of the source database as "clone1", create an additional instance on the destination client with the same SID as the clone database, that is, "clone1". This stores the clone instance properties. When you specify the clone options in the CommCell Console, the software pre-populates these properties.
- If you clone a database to an instance that resides on a proxy server, the following additional criteria must be met:
- The Oracle database version on the proxy server must be the same as the version on the source database.
- The Oracle user in both the source and proxy server must have the same user ID.
- The Oracle Agent and the MediaAgent must be installed on the proxy.
- The proxy server must have access to the storage array that contains the source database volumes.
- From the CommCell Browser, expand Client Computers > client > Oracle.
- Right-click the instance, point to All Tasks and then click Clone.
- On the Clone Options dialog box, select Latest Backup and click View Content.
- In the right pane of the CommCell Browser window, select the entire database to clone, and click Clone.
The Restore Options dialog box appears.
- On the General tab, specify the destination details.
- From the Destination Client list, select the clone destination.
- To use the RMAN Recovery catalog, clear the NO CATALOG check box, and then in the Catalog Connect boxes, type the connect string.
- On the Clone Options tab, specify the clone details.
- Specify the destination database.
Use a pre-defined instance
From the Instance list, select the destination.
Use a database that is not in the list
- In the Instance list, type the database name.
- In the ORACLE HOME box, type the location of the Oracle HOME directory.
- In the ORACLE User box, type the database user ID .
- Specify the location of the startup parameter file in the PFile box (for example, /oracle/clone/initdup1.ora).
If the file that you enter does not exist, the software creates the file and sets the parameters. The Oracle user must have writer permission for the startup file parameter directory.
If the file that you enter does exist, the dump and control files must exist and have writer permissions for the Oracle user. The specified server parameter file must meet the following criteria:
- The file must contain all the require Oracle initialization parameters
- The Oracle user must have permission to read the file
- Specify the path that will contain the Oracle database mounted snapshots, in the Snap Mount Location box (for example, /oracle/clone).
Note: The Oracle user must have write permissions to this directory. If you chose the option to create the server parameter file, the dump files are created in this directory, so there must be disk space to hold the logs until the reservation period expires.
- To have the software clean up any existing clone on the destination that has the same clone SID, regardless of the reservation period, select the Overwrite if clone exists check box.
- Specify time to retain the test environment in the Reservation Period box.
Note: The default time to retain the clone is 1 hour. When this time expires, the software dismounts the database and deletes the mount point.
- By default, the system sets up two redo log groups, with one log file in each group. To override this configuration (location, number of groups, files), select the Online Log Files check box, and then select the Group option.
- Click Add.
- In the Size box enter the size of the online redo log group member.
- Select the size specification of the member from the SizeSpec list. This value indicates that the value entered in the Size field should be interpreted as Kilobytes, Megabytes or Gigabytes.
- To allow the database to reuse an existing file, select the Reuse check box.
- Specify the destination database.
- On the Job Initiation tab, have the software automatically refresh the clone with the latest backup:
- Select Schedule.
- Click Configure Schedule Pattern.
For information on how to configure a restore schedule, see Schedule Restores.
- Optional: Recover the database. On the Recover tab, select the database recover option:
Recover to the current time
Enter the date and time.
Note: If the time zone is different on the CommServe computer and client, in the Time Zone box, select the client computer time zone.
The System Change Number (SCN) tracks the timing of transactions in the Oracle database. The SCNs are stored in the control files and the datafile headers. You can recover the database to the last existing SCN number in the control file, which is the last consistent database state.
Latest Backup Time
Recover to the latest backup job's completion time, which gets passed to the recover operation.
This value is the Completion Time on the General Tab of the Job Details dialog box. To view the value, right-click the job, and then select View Job Details. For instructions, see Job Controller - Getting Started.
Latest SCN Number
Recover to the latest SCN.
The latest backup job's Next SCN number is passed to the recover operation. This is the Next SCN on the General Tab of the Job Details dialog box. To view the value, right-click the job and then select View Job Details. For instructions, see Job Controller - Getting Started.
Relative to System Date
Recover to the number of days or hours relative to the current system date. The default option is to recover relative to the days.
If you select the Days option, the valid values are 0-7.
If you select the Hours option, the valid values are 0-23.
- Click OK.
Last modified: 1/21/2020 8:24:10 PM