RMAN Active Cloning From RAC TO RAC

• October 26, 2017 
•  Anuradha Mudgal
 
•  ORACLE 12C 
•  1 Comment

Being an Oracle DBA we know that what is cloning? This technical definition gives you a correct idea of the cloning. RMAN has the ability to duplicate, or clone, a database from a backup or from an active database. It is possible to create a duplicate database on a remote server with the same file structure, a remote server will a different file structure or the local server with a different file structure.

Below are the steps for doing RMAN from RAC to RAC. Steps will be same from NON RAC TO RAC .

Detail	Source	Destination
Node 1	prodb1-host	clonedb1-host
Node 2	prodb2-host	clonedb2-host
Version	12.1.0.2	12.1.0.2
Database Name	PRODB	CLONEDB
Instance Name 1	PRODB1	CLONEDB1
Instance Name 2	PRODB2	CLONEDB2

First we will restore the database to only one node1 and later we will convert to RAC.

All the activities need to be done target database only.

1. Create a static listener on the target host(NODE1)( From ORACLE_HOME)

LISTENER_CLONEDB =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = clonedb1-host)(PORT = 1527))
)

SID_LIST_LISTENER_CLONEDB =
(SID_LIST =
(SID_DESC =
(ORACLE_HOME = /oracle/app/oracle/product/12.1.0.2/dbhome_1)
(SID_NAME = CLONEDB1)
)
)

lsnrctl start LISTENER_CLONEDB

2. Add tns entry of both source and target db in tnsnames.ora file(TARGET_HOST NODE 1 )

 — SOUCE DB TNS
PRODB =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = prodb1-host)(PORT = 1521))
)
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = PRODB)
)
)– TARGET DB TNS
CLONEDB =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = clonedb1-host)(PORT = 1527))
)
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = CLONEDB)
)
)

3. Copy the pfile from source db and update the below parameters.

SOURCE PFILE LOOKS LIKE

*.cluster_database=TRUE
*.compatible=’12.1.0.2.0′
*.control_files=’+REDOA/PRODB/CONTROLFILE/current.256.911574607′,’+REDOB/PRODB/CONTROLFILE/current.256.911574607′,’+DATA/PRODB/CONTROLFILE/current.366.926770445’#Restore Controlfile
*.db_block_size=8192
*.db_domain=”
*.db_files=5000
*.db_name=’PRODB’
*.diagnostic_dest=’/oradbtrace’
PRODB1.instance_number=1
PRODB2.instance_number=2
*.large_pool_size=20g
*.log_archive_dest_1=’LOCATION=+ARCH’
*.log_archive_dest_state_1=’ENABLE’
*.open_cursors=4500
*.pga_aggregate_target=40G
*.processes=6000
*.remote_listener=’prodb-scan.stc.com.sa:1522′
*.remote_login_passwordfile=’exclusive’
*.sga_max_size=150G
*.sga_target=150G
*.shared_pool_size=21474836480
PRODB2.thread=2
PRODB1.thread=1
*.undo_retention=14400
PRODB2.undo_tablespace=’UNDOTBS1′
PRODB1.undo_tablespace=’UNDOTBS2′

Now change the db name to CLONEDB and instance names to CLONEDB1,CLONEDB2 in the pfile.
Apart from this, we need to do the few other changes. and save it as initCLONEDB1.ora in $ORACLE_HOME/dbs location.

*.log_archive_dest_1
*.db_create_file_dest
*.db_create_online_log_dest_1=’+REDOA’
*.control_files=’+REDOA’,’+REDOB’,’+DATA’
*.cluster_database=FALSE

Now the target db pfile will look as below.
Now change the db name to CLONEDB and instance names to CLONEDB1,CLONEDB2 in the pfile.
Apart from this, we need to do the few other changes. and save it as initCLONEDB1.ora in $ORACLE_HOME/dbs location.

*.log_archive_dest_1
*.db_create_file_dest
*.db_create_online_log_dest_1=’+REDOA’
*.control_files=’+REDOA’,’+REDOB’,’+DATA’
*.cluster_database=FALSE

Now the target db pfile will look as below.

cat initCLONEDB1.ora

*.cluster_database=FALSE
*.compatible=’12.1.0.2.0′
*.control_files=’+REDOA’,’+REDOB’,’+DATA’
*.db_block_size=8192
*.db_create_file_dest=’+DATA’
*.db_create_online_log_dest_1=’+REDOA’
*.db_create_online_log_dest_2=’+REDOB’
*.db_domain=”
*.db_files=5000
*.db_name=’CLONEDB’
*.diagnostic_dest=’/oradbtrace’
CLONEDB1.instance_number=1
CLONEDB2.instance_number=2
*.large_pool_size=20g
*.log_archive_dest_1=’LOCATION=+ARCH’
*.log_archive_dest_state_1=’ENABLE’
*.open_cursors=4500
*.pga_aggregate_target=40G
*.processes=6000
*.remote_listener=’clonedb-scan.stc.com.sa:1522′
*.remote_login_passwordfile=’exclusive’
*.sga_max_size=150G
*.sga_target=150G
*.shared_pool_size=21474836480
CLONEDB2.thread=2
CLONEDB1.thread=1
*.undo_retention=14400
CLONEDB2.undo_tablespace=’UNDOTBS1′
CLONEDB1.undo_tablespace=’UNDOTBS2′

4. Create password file on source and target db:(password should be same)

Now create password file on source and target db with same sys password.

— SOURCE DB ( PRODB)

cd $ORACLE_HOME/dbs
orapwd file=orapw$ORACLE_SID password=oracle force=y

— TARGET DB ( CLONEDB)

cd $ORACLE_HOME/dbs
orapwd file=orapw$ORACLE_SID password=oracle force=y

5. Now start the target database in nomount stage:

startup nomount pfile=initCLONEDB1.ora

6. Test the connectivity to target and auxiliary:[ TARGET HOST]

rman target sys/oracle#51234@PRODB auxiliary sys/oracle#51234@CLONEDB

Recovery Manager: Release 12.1.0.2.0 – Production on Tue Feb 21 19:52:37 2017
Copyright (c) 1982, 2014, Oracle and/or its affiliates. All rights reserved.

connected to target database: PRODB (DBID=678234706)
connected to auxiliary database: CLONEDB (not mounted)

For big databases, it is recommended to create a shell script and run in background as below.

7 . Prepare the rman run script:[TARGET HOST]

cat /export/home/oracle/rman_clone.cmd

run
{
allocate channel src1 type disk;
allocate channel src2 type disk;
allocate channel src3 type disk;
allocate channel src4 type disk;
allocate auxiliary channel aux1 type disk;
allocate auxiliary channel aux2 type disk;
allocate auxiliary channel aux3 type disk;
allocate auxiliary channel aux4 type disk;
allocate auxiliary channel aux5 type disk;
allocate auxiliary channel aux6 type disk;
duplicate target database to CLONEDB from active database USING BACKUPSET ;
}

8. Prepare the shell script.[TARGET HOST]

cat rman_clone_CLONEDB.sh
#!/bin/ksh
rman target sys/oracle#51234@PRODB auxiliary sys/oracle#51234@CLONEDB msglog /export/home/oracle/rman_clone2_log cmdfile=/export/home/oracle/rman_clone.cmd

9. Now run the script with nohup mode:[TARGET DB]

 nohup sh rman_clone_CLONEDB.sh &

10. Monitor the log

— FEW PART OF THE LOG

Recovery Manager: Release 12.1.0.2.0 – Production on Tue Feb 21 17:18:48 2017

Copyright (c) 1982, 2014, Oracle and/or its affiliates. All rights reserved.

connected to target database: PRODB (DBID=678234706)
connected to auxiliary database: CLONEDB (not mounted)

RMAN> run
2> {
3> allocate channel src1 type disk;
4> allocate channel src2 type disk;
5> allocate channel src3 type disk;
6> allocate channel src4 type disk;
7> allocate auxiliary channel aux1 type disk;
8> allocate auxiliary channel aux2 type disk;
9> allocate auxiliary channel aux3 type disk;
10> allocate auxiliary channel aux4 type disk;
11> allocate auxiliary channel aux5 type disk;
12> allocate auxiliary channel aux6 type disk;
13> duplicate target database to CLONEDB from active database USING BACKUPSET ;
14> }
15>
16>
using target database control file instead of recovery catalog
allocated channel: src1
channel src1: SID=2596 device type=DISK

allocated channel: src2
channel src2: SID=2847 device type=DISK

allocated channel: src3
channel src3: SID=3207 device type=DISK

allocated channel: src4
channel src4: SID=2539 device type=DISK

allocated channel: aux1
channel aux1: SID=5923 device type=DISK

allocated channel: aux2
.
.
.
.
.
.
.
.
.
.
input datafile copy RECID=196 STAMP=936563077 file name=+DATA/CLONEDB/DATAFILE/users.352.936559475
datafile 96 switched to datafile copy
input datafile copy RECID=197 STAMP=936563077 file name=+DATA/CLONEDB/DATAFILE/users.353.936559477
datafile 97 switched to datafile copy
input datafile copy RECID=198 STAMP=936563077 file name=+DATA/CLONEDB/DATAFILE/users.354.936559491
datafile 98 switched to datafile copy
input datafile copy RECID=199 STAMP=936563077 file name=+DATA/CLONEDB/DATAFILE/users.355.936559587
datafile 99 switched to datafile copy
input datafile copy RECID=200 STAMP=936563077 file name=+DATA/CLONEDB/DATAFILE/sysaux.356.936559587
datafile 100 switched to datafile copy
input datafile copy RECID=201 STAMP=936563077 file name=+DATA/CLONEDB/DATAFILE/system.357.936559595
datafile 101 switched to datafile copy
input datafile copy RECID=202 STAMP=936563077 file name=+DATA/CLONEDB/DATAFILE/undotbs1.358.936559689
datafile 102 switched to datafile copy
input datafile copy RECID=203 STAMP=936563077 file name=+DATA/CLONEDB/DATAFILE/undotbs2.359.936559925
datafile 103 switched to datafile copy
input datafile copy RECID=204 STAMP=936563077 file name=+DATA/CLONEDB/DATAFILE/users.360.936560101
Reenabling controlfile options for auxiliary database
Executing: alter database add supplemental log data
Executing: alter database add supplemental log data(PRIMARY KEY, UNIQUE) columns
Executing: alter database enable block change tracking

contents of Memory Script:
{
Alter clone database open resetlogs;
}
executing Memory Script

database opened
Cannot remove created server parameter file
Finished Duplicate Db at 21-FEB-17
released channel: src1
released channel: src2
released channel: src3
released channel: src4

Complete log:

Now the database is up and available but only on one node. We need to register the database and convert it into a RAC database.

11. Make cluster parameter to true:

alter system set cluster_database=TRUE scope=spfile sid=’*’;

12. Create spfile in asm disk

create pfile=’/export/home/oracle/test.ora’ from spfile;

— Now create the spfile in ASM dis

create spfile=’+DATA/CLONEDB/PARAMETERFILE/spfileCLONEDB.ora’ from pfile;

Go to $ORACLE_HOME/dbs

Update initCLONEDB1.ora as below

cat initCLONEDB1.ora
SPFILE=’+DATA/CLONEDB/PARAMETERFILE/spfileCLONEDB.ora’

Same thing do it on node 2:(upgrade initCLONEDB2.ora)

cat initCLONEDB2.ora
SPFILE=’+DATA/CLONEDB/PARAMETERFILE/spfileCLONEDB.ora’

13. Register the database and modify the configuration

 srvctl add database -db CLONEDB -oraclehome /oracle/app/oracle/product/12.1.0.2/dbhome_1
srvctl add instance -d CLONEDB -i CLONEDB1 -node clonedb1-host
srvctl add instance -d CLONEDB -i CLONEDB2 -node clonedb1-host
srvctl modify database -d CLONEDB -spfile +DATA/CLONEDB/PARAMETERFILE/spfileCLONEDB.orasrvctl config database -d CLONEDB
14. Now shutdown the database on node 1 and restart using srvctl.

 shutdown immediate;

srvctl start database -d CLONEDB

— Check crs status

crsctl stat res -t

ora.clonedb.db
1 ONLINE ONLINE sec54-1 Open,STABLE
2 ONLINE ONLINE sec55-1 Open,STABLE

Now RAC database is ready for use.

RESTORE FAILED RMAN ACTIVE CLONE:

While restoring a big database, cloning might fail in the mid due to a network issue. In that case, no need to start from the beginning.
Follow below steps to resume the restore.It will skip the copied files,

1. Shutdown the database
2. Remove the spfile, which rman has created during the duplication
3. Start the database with nomount stage(using the original init file, which you have used , which starting the database initially)
4. Run the same exact rman duplication script.
5. Rman will detect the files which were copied during the previous cloning process and those will be skipped.It will only restore the new datafile which was not copied previously.

It will show the warning like below:
Using previous duplicated file +DATA/CLONEDB/DATAFILE/undotbs2.628.936641735 for datafile 78 with checkpoint SCN of 11712137096850. You can ignore.

NOTE :

If the database version is 11g, then use the below rman run block:(step 7 )

run
{
allocate channel src1 type disk;
allocate channel src2 type disk;
allocate channel src3 type disk;
allocate channel src4 type disk;
allocate auxiliary channel aux1 type disk;
allocate auxiliary channel aux2 type disk;
allocate auxiliary channel aux3 type disk;
allocate auxiliary channel aux4 type disk;
allocate auxiliary channel aux5 type disk;
allocate auxiliary channel aux6 type disk;
duplicate target database to CLONEDB from active database ;
}