Cloning an Oracle/DB2/Sybase Database in a SAN Environment Using EVA Snapclone Technology (August 2007)

14
generated after the Snapclone is created, as a result of the volume clone, copy it to
the archived logs directory.
9. Recover the clone TEST. Enter the following command on the clone instance:
SQL> recover database using backup controlfile;
You are prompted to enter the archived log file names one by one for recovery. You
can select the AUTO option. When the archived logs are recovered, select the
filename option and enter the names of the previously copied files, redo03a.log and
redo3b.log. The following message is displayed:
Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
/prod/clone/archlog/redo03b.log
Log applied.
Media recovery complete.
When the recovery process for the clone TEST is complete, open the clone TEST with
the resetlogs option by entering the following command:
sql> alter database open resetlogs;
Database altered.
The clone TEST is now ready for use. HP recommends that you back up the clone
immediately, because the archived redo logfiles are now reset. The archived redo
logfiles cannot be used to recreate the clone. The above database cloning procedures
are shown step by step for clarification. In general, all the steps can be automated in
a single script. For a summary of the above steps, see
Quick Reference Steps for Cloning
the Oracle Database.
Cloning a Full Database Using Cold Backup
The previous section described how to clone a production database using hot backup, in
which the database is open. In a cold backup, the database is shut down and no activity is
in progress. As a result, the database is stable.
From the EVA point of view, there is no difference between a hot backup and a cold backup.
The Snapclones are only the EVA LUNs that constitute the database. Therefore, the steps to
create a cloned database discussed for a hot backup are still valid for a cold backup, except
that the SQL commands, with BEGIN BACKUP and END BACKUP, are not necessary.