Oracle RAC恢復成單節點資料庫

In this Document

APPLIES TO:

GOAL

- You have a RAC database backed up by RMAN to disk location
- You need to restore this backup as Single Instance on another node

SOLUTION

1) Take appropriate RMAN backup of the production RAC database. Note that you should turn on the CONTROLFILE AUTOBACKUP configuration so that we have the controlfile backed up after the database backup. When we restore the controlfile on new host from this autobackup piece, it will have the information of the latest backup.

RMAN> run{
2> allocate channel c1 type disk format '/oracle/10g/backup/%U';
3> backup database;
4> backup archivelog all;
5> }

allocated channel: c1
channel c1: sid=133 instance=racdb1 devtype=DISK

Starting backup at 12-FEB-07
channel c1: starting full datafile backupset
channel c1: specifying datafile(s) in backupset
input datafile fno=00001 name=/ocfs2/oradata/racdb/system01.dbf
input datafile fno=00002 name=/ocfs2/oradata/racdb/undotbs01.dbf
input datafile fno=00005 name=/ocfs2/oradata/racdb/undotbs02.dbf
input datafile fno=00003 name=/ocfs2/oradata/racdb/sysaux01.dbf
input datafile fno=00004 name=/ocfs2/oradata/racdb/users01.dbf
channel c1: starting piece 1 at 12-FEB-07
channel c1: finished piece 1 at 12-FEB-07
piece handle=/oracle/10g/backup/09i9sruq_1_1 tag=TAG20070212T162458 comment=NONE
channel c1: backup set complete, elapsed time: 00:01:05
Finished backup at 12-FEB-07

Starting backup at 12-FEB-07
current log archived
channel c1: starting archive log backupset
channel c1: specifying archive log(s) in backup set
input archive log thread=1 sequence=56 recid=3 stamp=613417105
input archive log thread=1 sequence=57 recid=4 stamp=613417106
input archive log thread=1 sequence=58 recid=5 stamp=614363168
input archive log thread=2 sequence=1 recid=1 stamp=613417090
input archive log thread=2 sequence=2 recid=2 stamp=613417093
input archive log thread=2 sequence=3 recid=6 stamp=614363170
channel c1: starting piece 1 at 12-FEB-07
channel c1: finished piece 1 at 12-FEB-07
piece handle=/oracle/10g/backup/0ai9ss14_1_1 tag=TAG20070212T162610 comment=NONE
channel c1: backup set complete, elapsed time: 00:00:04
Finished backup at 12-FEB-07

Starting Control File and SPFILE Autobackup at 12-FEB-07
piece handle=/oracle/10g/backup/c-610677177-20070212-00 comment=NONE
Finished Control File and SPFILE Autobackup at 12-FEB-07
released channel: c1

2) Create a PFILE for the single instance database using the production RAC parameter file
a) don't forget to modify the following parameters depending on the directory structure of the new host: audit_file_dest, background_dump_dest, control_files, core_dump_dest, log_archive_dest_1, user_dump_dest etc
b) remove RAC specific parameters such as cluster_database_instances, cluster_database etc
c) for the parameter undo_tablespace, mention any one undo tablespace name

Known issue which may require an additional parameter to be set :
   Note 334899.1 RMAN Duplicate from RAC backup fails ORA-38856

3) Move the backup pieces and the modified INIT.ORA file to the new host. Starting from 10g it is NO longer compulsory to copy the RMAN backup pieces to exactly the same location on the new host as the production location.

4) Use the pfile created above to STARTUP NOMOUNT the database on the new host

oracle@test-br ractest]$ sqlplus "/ as sysdba"

SQL*Plus: Release 10.2.0.1.0 - Production on Fri Feb 16 03:14:23 2007

Copyright (c) 1982, 2005, Oracle.  All rights reserved.

Connected to an idle instance.

SQL> startup nomount;
ORACLE instance started.

Total System Global Area  167772160 bytes
Fixed Size                  1218292 bytes
Variable Size              62916876 bytes
Database Buffers           96468992 bytes
Redo Buffers                7168000 bytes

5) Now invoke RMAN and restore the controlfile specifying the location where the controlfile autobackup piece is restored on this new server. You can mount the database once the controlfile is restored successfully.

[oracle@test-br ractest]$ rman target / nocatalog

Recovery Manager: Release 10.2.0.1.0 - Production on Fri Feb 16 03:16:31 2007

Copyright (c) 1982, 2005, Oracle.  All rights reserved.

connected to target database: racdb (not mounted)
using target database control file instead of recovery catalog

RMAN> restore controlfile from '/u01/oracle/oradata/ractest/c-610677177-20070212-00';

Starting restore at 16-FEB-07
allocated channel: ORA_DISK_1
channel ORA_DISK_1: sid=156 devtype=DISK

channel ORA_DISK_1: restoring control file
channel ORA_DISK_1: restore complete, elapsed time: 00:00:02
output filename=/u01/oracle/oradata/ractest/control01.ctl
Finished restore at 16-FEB-07

RMAN> alter database mount;

database mounted
released channel: ORA_DISK_1

6) You can skip this step if you have restored the RMAN backup pieces to exactly the same location they were backed up on production. If this is not the case then you need to catalog the RMAN backup pieces to make RMAN aware of thier new location on the new host. Note that CATALOG BACKUPPIECE command is available only starting from 10g.

RMAN> catalog backuppiece '/u01/oracle/oradata/ractest/09i9sruq_1_1';

cataloged backuppiece
backup piece handle=/u01/oracle/oradata/ractest/09i9sruq_1_1 recid=10 stamp=614661579

RMAN> catalog backuppiece '/u01/oracle/oradata/ractest/0ai9ss14_1_1';

cataloged backuppiece
backup piece handle=/u01/oracle/oradata/ractest/0ai9ss14_1_1 recid=11 stamp=614661599

7) Now we'll determine the point upto which media recovery should run on the restored database.

RMAN> list backup of archivelog all;

List of Archived Logs in backup set 9
Thrd Seq     Low SCN    Low Time  Next SCN   Next Time
---- ------- ---------- --------- ---------- ---------
1    56      214541     01-FEB-07 226238     01-FEB-07
1    57      226238     01-FEB-07 226240     01-FEB-07
1    58      226240     01-FEB-07 233107     12-FEB-07
2    1       186185     28-JAN-07 225714     01-FEB-07
2    2       225714     01-FEB-07 226037     01-FEB-07
2    3       226037     01-FEB-07 233110     12-FEB-07

Check the last archive sequence for all redo threads and select the archive sequence having LEAST "Next SCN" among them. In our case sequence 58 of thread 1 has Next SCN of 233107 while sequence 3 of thread 2 has Next SCN of 233110. Since squence 58 of thread 1 has least Next SCN we will recover upto this point. (If you are keen to have recovery run until some specific time you can always give SET UNTIL TIME)

8) Having determined the point upto which media recovery should run, start the restore/recovery using:

RMAN> run {
2> set until sequence 59 thread 1;
3> set newname for datafile 1 to '/u01/oracle/oradata/ractest/data/system01.dbf';
4> set newname for datafile 2 to '/u01/oracle/oradata/ractest/data/undotbs01.dbf';
5> set newname for datafile 3 to '/u01/oracle/oradata/ractest/data/sysaux01.dbf';
6> set newname for datafile 4 to '/u01/oracle/oradata/ractest/data/users01.dbf';
7> set newname for datafile 5 to '/u01/oracle/oradata/ractest/data/undotbs02.dbf';
8> restore database;
9> switch datafile all;
10> recover database;
11> }

Since we determined previously that media recovery should run until sequence 58 hence we use  SET UNTIL SEQUENCE 59 (+1) above. You also need to use SET NEWNAME clause to restore datafiles to a location on the new host which is different from the production path. Finally, SWITCH DATAFILE ALL clause updates these new datafile locations in the controlfile.

9) Once RMAN restore/recovery finishes, you will want to rename the online redolog files before opening the database in case the production path of redo log files is not available on the new host. After renaming the redolog files, the database can be opened with RESETLOGS

SQL> select member from v$logfile;

MEMBER
--------------------------------------------------------------------------------
/ocfs2/oradata/racdb/redo01.log
/ocfs2/oradata/racdb/redo02.log
/ocfs2/oradata/racdb/redo03.log
/ocfs2/oradata/racdb/redo04.log
/ocfs2/oradata/racdb/redo05.log
/ocfs2/oradata/racdb/redo06.log

6 rows selected.

SQL> alter database rename file '/ocfs2/oradata/racdb/redo01.log' to '/u01/oracle/oradata/ractest/log/redo01.log';

Database altered.

SQL> alter database rename file '/ocfs2/oradata/racdb/redo02.log' to '/u01/oracle/oradata/ractest/log/redo02.log';

Database altered.

SQL> alter database rename file '/ocfs2/oradata/racdb/redo03.log' to '/u01/oracle/oradata/ractest/log/redo03.log';

Database altered.

SQL> alter database rename file '/ocfs2/oradata/racdb/redo04.log' to '/u01/oracle/oradata/ractest/log/redo04.log';

Database altered.

SQL> alter database rename file '/ocfs2/oradata/racdb/redo05.log' to '/u01/oracle/oradata/ractest/log/redo05.log';

Database altered.

SQL> alter database rename file '/ocfs2/oradata/racdb/redo06.log' to '/u01/oracle/oradata/ractest/log/redo06.log';

Database altered.

SQL> alter database open resetlogs;

Database altered.

10) Once the database is opened successfully, you may remove the redolog groups for redo threads of other instances

SQL> select THREAD#, STATUS, ENABLED
  2  from v$thread;

   THREAD# STATUS ENABLED
---------- ------ --------
         1 OPEN   PUBLIC
         2 CLOSED PRIVATE

SQL> select group# from v$log where THREAD#=2;

    GROUP#
----------
         4
         5
         6

SQL> alter database disable thread 2;

Database altered.

SQL> alter database drop logfile group 4;
alter database drop logfile group 4
*
ERROR at line 1:
ORA-00350: log 4 of instance racdb2 (thread 2) needs to be archived
ORA-00312: online log 4 thread 2: '/u01/oracle/oradata/ractest/log/redo04.log'

SQL> alter database clear unarchived logfile group 4;

Database altered.

SQL> alter database drop logfile group 4;

Database altered.

SQL> alter database drop logfile group 5;

Database altered.

SQL> alter database drop logfile group 6;

Database altered.

SQL> select THREAD#, STATUS, ENABLED from v$thread;

   THREAD# STATUS ENABLED
---------- ------ --------
         1 OPEN   PUBLIC

11) Now you can remove the undo tablespaces of other instances and create a new temporary tablespace to complete the activity.

SQL> sho parameter undo;

NAME                                 TYPE        VALUE
------------------------------------ ----------- ------------------------------
undo_management                      string      AUTO
undo_retention                       integer     900
undo_tablespace                      string      UNDOTBS1
SQL>
SQL>
SQL> select tablespace_name from dba_tablespaces where contents='UNDO';

TABLESPACE_NAME
------------------------------
UNDOTBS1
UNDOTBS2

SQL> drop tablespace UNDOTBS2 including contents and datafiles;

Tablespace dropped.

SQL> select name from v$tempfile;

NAME
--------------------------------------------------------------------------------
/ocfs2/oradata/racdb/temp01.dbf


SQL> select tablespace_name from dba_tablespaces where contents='TEMPORARY';

TABLESPACE_NAME
------------------------------
TEMP

SQL> create temporary tablespace TEMP1
  2  tempfile '/u01/oracle/oradata/ractest/data/temp01.dbf'
  3  size 50M;

Tablespace created.

SQL> alter database default temporary tablespace TEMP1;

Database altered.

SQL> drop tablespace TEMP including contents and datafiles;

Tablespace dropped.

REFERENCES