RMAN Restore and Recovery Scenarios

Full Database Restore

$ORACLE_HOME/bin/rman target / nocatalog
RMAN> shutdown abort;
RMAN> startup mount;
RMAN> restore database;
RMAN> recover database;
RMAN> alter database open;

 

Tablespace Restore (online)

$rman target / nocatalog
RMAN> sql ‘alter tablespace users offline';
RMAN> restore tablespace users;
RMAN> recover tablespace users;
RMAN> sql ‘alter tablespace users online';

 

Tablespace Restore (offline)

$rman target / nocatalog
RMAN> shutdown abort;
RMAN> startup mount;
RMAN> restore tablespace users;
RMAN> recover tablespace users;
RMAN> alter database open;
database opened

 

Restoring a Specific Datafile

$rman target / nocatalog
RMAN> shutdown abort;
RMAN> startup mount;
RMAN> restore datafile ‘/oradata/DB1/dbf/users01.dbf';
RMAN> recover datafile ‘/oradata/DB1/dbf/users01.dbf';
RMAN> alter database open;
database opened

 

Control File Restoration

Prerequisite: In your rman backup directory determine the latest control file backup.
Default Format: c-nnnnnnnnnn-nnnnnnnn-nn

$rman target / nocatalog
RMAN> shutdown abort;
RMAN> startup nomount;
RMAN> set dbid = 1184749195
RMAN> restore controlfile from ‘/oradata/DB1/rman/c-1184749195-20060626-02′
RMAN> alter database mount;
RMAN> restore database;
RMAN> recover database;
RMAN> alter database open resetlogs;
database opened
Database Point-In-Time-Recovery (PITR)

Also known as time-based incomplete recovery.

$rman target / nocatalog
RMAN> shutdown abort;
RMAN> startup mount;
RMAN> restore database until time “to_date(’09/03/07 13:00:00′, ‘MM/DD/YY HH24:MI:SS’)”;
RMAN> recover database until time “to_date(’09/03/07 13:00:00′, ‘MM/DD/YY HH24:MI:SS’)”;
RMAN> alter database open resetlogs;
database opened

* Make sure you perform a full backup after this operation!

Restore to Another System

Prerequisites
Ideally ensure destination system configured exactly like source.
Same OS version and patch level.
Same drives (C:, D:, S: etc.).
CPU and RAM same or better.
The same version of Oracle is installed on the target system as the source.
Ensure the ORACLE_HOME and ORACLE_SID environment variables are set.
Ensure the listener is running.

Copy RMAN backupset files to the destination system rman directory.

Procedure

Restore SPFILE and Control File

$rman target / nocatalog
RMAN> set dbid 161080442
RMAN> startup nomount;
Creates the file: %ORACLE_HOME%\database\hc_db1.dat
RMAN> restore spfile from ‘R:\rman\C-161080442-20080313-00′;
Creates the file: %ORACLE_HOME%\database\SPFILEDB1.ORA
RMAN> startup force nomount
RMAN> restore controlfile from ‘R:\rman\C-161080442-20080313-00′;
RMAN> shutdown immediate
RMAN> exit


Restore and Recover the Data

$rman target / nocatalog
RMAN> startup mount;
RMAN> restore database;
For a large database this step may take some time.
RMAN> recover database;
If you do not have\need the very last log(s) you can disregard any error messages.
ORA-00310: archived log contains sequence 100; sequence 101 required…
RMAN> alter database open resetlogs;
database opened

Scenario Basis of Recovery

1-  If you lost all data files


SQL> startup mount;
RMAN> restore database;
RMAN> recover database;
SQL> alter database open;

2- If you lost a tablespace


SQL> alter tablespace users offline;
RMAN> restore tablespace users;
RMAN> recover tablespace users;
SQL> alter tablespace users online;

 if you can not offline tablespace;


$ sqlplus “/ as sysdba”
SQL> shutdown abort;
SQL> startup mount;
$ rman target / catalog_user/catalog_user_password@catalogdb
RMAN> restore tablespace users;
RMAN> recover tablespace users;
SQL> alter database open;

3- if you lost a datafile


SQL> alter database datafile '/oracle/oradata/users.dbf' offline;
RMAN> restore datafile '/oracle/oradata/users.dbf'
RMAN> recover datafile '/oracle/oradata/users.dbf'
SQL> alter database datafile '/oracle/oradata/users.dbf' online;
 
if you cannot offline datafile;

$ sqlplus “/ as sysdba”
SQL> shutdown abort;
SQL> startup mount;
$ rman target / catalog_user/catalog_user_password@catalogdb
RMAN> restore datafile '/oracle/oradata/users.dbf';
RMAN> recover datafile '/oracle/oradata/users.dbf';
SQL> alter database open;
 


4-  if you lost your controlfiles


$ sqlplus “/ as sysdba”
SQL> shutdown abort;
SQL> startup nomount;
$ rman target / catalog_user/catalog_user_password@catalogdb
RMAN> set dbid = 3970640872;
RMAN> restore controlfile;
SQL> alter database mount;
SQL> alter database open;

you will receive an error ORA-01589 when you open database
 ORA-01589: must use RESETLOGS or NORESETLOGS option for database open


SQL> shutdown abort;
SQL> startup mount;
$ rman target / catalog_user/catalog_user_password@catalogdb
RMAN> recover database;
SQL> alter database open resetlogs;
RMAN> reset database;

if you open database with resetlogs, SCN number will be zero. In this situation
 all previous backups will be invalid. You must full backup.

5- May be a special situation. You need to incomplete recovery


 A. Time-Based incomplete recovery


$ sqlplus "/ as sysdba"
SQL> shutdown abort;
SQL> startup mount;
$ rman target / catalog_user/catalog_user_password@catalogdb
RMAN> restore database until time "to_date('06/05/11 12:0:00','DD/MM/YY HH24:MI:SS')";
RMAN> recover database until time "to_date('06/05/11 12:0:00','DD/MM/YY HH24:MI:SS')";
SQL> alter database open resetlogs;
 
 

 B. SCN-Based incomplete recovery


$ sqlplus "/ as sysdba"
SQL> shutdown abort;
SQL> startup mount;
$ rman target / catalog_user/catalog_user_password@catalogdb
RMAN> restore database until scn 1000;
RMAN> recover database until scn 1000;
SQL> alter database open resetlogs;


 C. Archive log sequence based incomplete recovery


$ sqlplus "/ as sysdba"
SQL> shutdown abort;
SQL> startup mount;
$ rman target / catalog_user/catalog_user_password@catalogdb
RMAN> restore database until sequence 9923;
RMAN> recover database until sequence 9923;
SQL> alter database open resetlogs;


6-  if you need some archive logs in your backup


$ rman target / catalog_user/catalog_user_password@catalogdb
RMAN> restore ARCHIVELOG FROM TIME 'SYSDATE-1' UNTIL TIME 'SYSDATE';
 
OR


RMAN> restore ARCHIVELOG FROM TIME "to_date('07/11/05 00:00:01','MM/DD/YY HH24:MI:SS')
UNTIL TIME 'SYSDATE';
 

7- if your data block is corrupted you will receive an error below.

Error:
 ORA-01578: ORACLE data block corrupted (file # 8, block # 13)
 ORA-01110: data file 8: ‘/oracle/oradata/users.dbf’

for recover data block;


$ rman target / catalog_user/catalog_user_password@catalogdb
RMAN>blockrecover datafile 8 block 13;
 
For Block-Level Media Recovery – Concept & Example (Doc ID 144911.1)

 To recover, we can give a specific backup set;


# recovery from backupset

RMAN> BLOCKRECOVER DATAFILE 8 BLOCK 13 DATAFILE 2 BLOCK 19 FROM BACKUPSET;

# recovery from image copy

RMAN> BLOCKRECOVER DATAFILE 8 BLOCK 13 DATAFILE 2 BLOCK 19
      FROM DATAFILECOPY;

# recovery from backupset which have "FULL" tag

RMAN> BLOCKRECOVER DATAFILE 8 BLOCK 13 DATAFILE 2 BLOCK 199
       FROM TAG = FULL;
During backup or “Validate Backup” command, RMAN finds corrupted blocks and writes to V$DATABASE_BLOCK_CORRUPTION view. When the RMAN recover the corrupt block then automatically updates this view. List of all the corruption of the past, can be viewed over V$BACKUP_CORRUPTION and V$COPY_CORRUPTION views.  If you run the following command,  RMAN will recover all the corrupted blocks in view V$DATABASE_BLOCK_CORRUPTION.


  RMAN>BLOCKRECOVER CORRUPTION LIST RESTORE UNTIL TIME 'SYSDATE-10';

8- if you have a image copy backup and your datafile number 2 has problems then you

can switch datafile number2 to image copy.


RMAN>sql ‘alter database datafile 2 offline’;
RMAN>switch datafile 2 to copy;
RMAN>recover datafile 2;
RMAN>sql ‘alter database datafile 2 online’;

  

RAC : 11gR2 Clusterware Startup Sequence 

Here is the brief explanation that how the clusterware brings up step by step .

1. When a node of an Oracle Clusterware cluster start/restarts, OHASD is started by platform-specific means. OHASD is the root for bringing up Oracle Clusterware. OHASD has access to the OLR (Oracle Local Registry) stored on the local file system. OLR provides needed data to complete OHASD initialization.

2. OHASD brings up GPNPD and CSSD. CSSD has access to the GPNP Profile stored on the local file system. This profile contains the following vital bootstrap data;

        a. ASM Diskgroup Discovery String

        b. ASM SPFILE location (Diskgroup name)

        c. Name of the ASM Diskgroup containing the Voting Files

3. The Voting Files locations on ASM Disks are accessed by CSSD with well-known pointers in the ASM Disk headers and CSSD is able to complete initialization and start or join an existing cluster.

4. OHASD starts an ASM instance and ASM can now operate with CSSD initialized and operating. The ASM instance uses special code to locate the contents of the ASM SPFILE, assuming it is stored in a Diskgroup.

5. With an ASM instance operating and its Diskgroups mounted, access to Clusterware’s OCR is available to CRSD.

6. OHASD starts CRSD with access to the OCR in an ASM Diskgroup.

7. Clusterware completes initialization and brings up other services under its control.

When Clusterware starts three files are involved.

1. OLR – Is the first file to be read and opened. This file is local and this file contains information regarding where the voting disk is stored
and information to startup the ASM. (e.g ASM DiscoveryString)

2. VOTING DISK – This is the second file to be opened and read, this is dependent on only OLR being accessible.

ASM starts after CSSD or ASM does not start if CSSD is offline (i.e voting file missing)

How are Voting Disks stored in ASM?

Voting disks are placed directly on ASMDISK. Oracle Clusterware will store the votedisk on the disk within a disk group that holds the Voting Files.
Oracle Clusterware does not rely on ASM to access the Voting Files, which means Oracle Clusterware does not need of Diskgroup to read and write on ASMDISK. It is possible to check for existence of voting files on a ASMDISK using the V$ASM_DISK column VOTING_FILE.
So, voting files not depend of Diskgroup to be accessed, does not mean that the diskgroup is not needed, diskgroup and voting file are linked by their settings.

3. OCR – Finally the ASM Instance starts and mount all Diskgroups, then Clusterware Deamon (CRSD) opens and reads the OCR which is stored on Diskgroup.

So, if ASM already started, ASM does not depend on OCR or OLR to be online. ASM depends on CSSD (Votedisk) to be online.

There is a exclusive mode to start ASM without CSSD (but it’s to restore OCR or VOTE purposes)