RMAN Backup Restoration Drill

This post explains the procedure to test the RMAN backups on a test box. Here I am doing only partial restoration (not full database) due to space constraints.

1. Copy the pfile from prod box to test box and edit it.
$ scp init.ora oracle@TestBox:/restore_path

2. Create password file
$ orapwd file=orapw password=oracle

3. Create necessary folders in /restore_path
$ cd /restore_path
$ mkdir adump bdump cdump udump redo data archive

4. Restore the control file
$ export ORACLE_SID=
$ sqlplus "/ as sysdba"
sql> startup nomount pfile=/restore_path/init.ora
sql> exit
$ rman target /
rman> run {
allocate channel c1 type disk;
restore controlfile from '/restore_path/rman_controlfile_bkp.ctl';
}

5. Catalog all the backup files
rman> alter database mount;
rman> catalog backuppiece '/restore_path/ORADB_rman_full_bkp_s5_p1;
rman> catalog backuppiece '/restore_path/ORADB_rman_full_bkp_s4_p2;
rman> catalog backuppiece '/restore_path/ORADB_rman_full_bkp_s3_p3;
...

6. Restore selected tablespace
rman> run {
allocate channel c1 type disk;
allocate channel c2 type disk;
allocate channel c3 type disk;
allocate channel c4 type disk;
allocate channel c5 type disk;
set newname for datafile 1 to '/restore_path/data/system01.dbf';
set newname for datafile 3 to '/restore_path/data/sysaux01.dbf';
set newname for datafile 42 to '/restore_path/data/audit_data01.dbf';
set newname for datafile 2 to '/restore_path/data/audit_data.dbf';
set newname for datafile 10 to '/restore_path/data/undotbs01.dbf';
restore tablespace system;
restore tablespace syasaux;
restore tablespace undotbs01;
restore tablespace audit_data;
switch datafile all;
}

7. Drop all the remaining datafiles which are not included in restore drill
alter database datafile 5 offline off;
alter database datafile 15 offline off;
alter database datafile 25 offline off;
alter database datafile 35 offline off;
.....

8. Recover database
$ export ORACLE_SID=
$ sqlplus "/ as sysdba"
sql> recover database using backup contrlfile until cancel;

9. Restore archive logs
$ rman target /
rman> run {
set archivelog destination to '/restore_path/archive/';
allocate channel c1 type disk;
allocate channel c2 type disk;
restore archivelog from sequence 2503 until sequence 2150;
}

9.1 For RAC
restore archivelog from sequence 2503 until sequence 2150 thread=1;

restore archivelog from sequence 2503 until sequence 2150 thread=2;

10. Rename Redo log files
sql> alter database rename file '/prod_path/redo/reado01.log' to
'/restore_path/redo/reado01.log';

11. Cleare all redo log groups
sql> alter database cleare logfile group 1;
sql> alter database cleare logfile group 2;
sql> alter database cleare logfile group 3;
.....

12. Open the database with openrestlog option
sql> alter database open resetlogs;

13. Query the database
sql> select name,open_mode,log_mode,database_role from v$database;

14. Drop the database
sql> shutdown abort;
sql> startup nomount;
sql> alter database mount exclusive;
sql> alter system enable restricted session;
sql> drop database;

15. Drop all the directories created for restoration
$ rmdir adump bdump cdump udump redo data archive