Step by step oracle 10GR2 data guard installation with RMAN duplicate.

Notes: 

1. Virtualbox on Ubuntu 11.10 with 2 virtual machine  Oracle Linux 5 is used for Data Guard installation.
2. First installed Oracle Linux 5 on Virtualbox (hostname=oracle10g), installed requied rpms, installed oracle software and a database named ORCL. Then clone this virtual machine as hostname=oracle10gs
3. Primary hostname : oracle10g, Standby hostname: oracle10gs
4. Primary database UNIQUE NAME: ORCL, standby database UNIQUE NAME: ORCL_STBY
5. Primarydatabase Net Service Name: ORCL, standby database Net Service Name: ORCL_STBY
6. This is a phiysical standby database with Maximum Performance mode. In this mode: Transactions on the primary commit as soon as redo information has been written to the online redo log. Transfer of redo information to the standby server is asynchronous, so it does not impact on performance of the primary.

ON PRIMARY SIDE:

• Check for Primary database whether archivelog mode is on.

SQL> SELECT log_mode FROM v$database;

LOG_MODE
------------
NOARCHIVELOG

Put the database to archivelog mode:

SQL> SHUTDOWN IMMEDIATE;
SQL> STARTUP MOUNT;
SQL> ALTER DATABASE ARCHIVELOG;
SQL> ALTER DATABASE OPEN;

• Enable force logging:

SQL> ALTER DATABASE FORCE LOGGING;

• Create password file if it is not exist:

$ orapwd file=$ORACLE_HOME/dbs/orapwORCL password=****** entries=30

• Add these parameters to primary instance ORCL init file:

DB_UNIQUE_NAME=ORCL 

LOG_ARCHIVE_CONFIG='DG_CONFIG=(ORCL,ORCL_STBY)'
LOG_ARCHIVE_DEST_2='SERVICE=ORCL_STBY NOAFFIRM ASYNC VALID_FOR=(ONLINE_LOGFILES,PRIMARY_ROLE) DB_UNIQUE_NAME=ORCL_STBY' 

LOG_ARCHIVE_DEST_STATE_2=ENABLE

REMOTE_LOGIN_PASSWORDFILE=EXCLUSIVE 

LOG_ARCHIVE_FORMAT='%t_%s_%r.arc'

LOG_ARCHIVE_MAX_PROCESSES=30

STANDBY_FILE_MANAGEMENT=AUTO

STANDBY_ARCHIVE_DEST='/u01/db_files/ARCH'

• Primary Database: Standby Role Initialization Parameters

These parameters are needed when there is a switchower between primary role and standby role. 

###################################################

#These parameters are needed when switchover occurs.

###################################################

#FAL_SERVER=ORCL_STBY

#FAL_CLIENT=ORCL

######################

#If file name conversion is needed:

#DB_FILE_NAME_CONVERT='boston','chicago'

#LOG_FILE_NAME_CONVERT='/arch1/boston/','/arch1/chicago/'

######################

#STANDBY_FILE_MANAGEMENT=AUTO

#############################################

• Modify tnsnames.ora by adding standby instance and catalog database instance of RMAN:

ORCL_STBY =

  (DESCRIPTION =

    (ADDRESS = (PROTOCOL = TCP)(HOST = oracle10gs)(PORT = 1521))

    (CONNECT_DATA =

      (SERVER = DEDICATED)

      (SERVICE_NAME = ORCL)

    )

  )

CATALOG =

  (DESCRIPTION =

    (ADDRESS = (PROTOCOL = TCP)(HOST = oracle10g)(PORT = 1522))

    (CONNECT_DATA =

      (SERVER = DEDICATED)

      (SERVICE_NAME = CATALOG)

    )

  )

• Add standby logfiles: 

Number of standby logfile groups must be +1 of number of redo log file groups. In my case, number of redolog file groups:3. I need to create 4 standby redolog file groups.

ALTER DATABASE ADD STANDBY LOGFILE GROUP 5 ('/u01/oradata/standby05.dbf') SIZE 50M

ALTER DATABASE ADD STANDBY LOGFILE GROUP 6 ('/u01/oradata/standby06.dbf') SIZE 50M

ALTER DATABASE ADD STANDBY LOGFILE GROUP 7 ('/u01/oradata/standby07.dbf') SIZE 50M

ALTER DATABASE ADD STANDBY LOGFILE GROUP 8 ('/u01/oradata/standby08.dbf') SIZE 50M

• Backup the database  using RMAN  with standby controlfile

connect to rman:

rman catalog rman/rman@CATALOG target /

Run the script:

run

{

allocate channel ch1 type disk format '/u01/backup/std_%U';

allocate channel ch2 type disk format '/u01/backup/std_%U';

BACKUP DATABASE INCLUDE CURRENT CONTROLFILE FOR STANDBY PLUS ARCHIVELOG;

}

• Copy or share these backup files under /u01/backup to standby instance

$ scp /u01/backup/* oracle@oracle10gs:/u01/backup

• Copy the init file to standby side

$ scp  $ORACLE_HOME/dbs/initORCL.ora oracle@oracle10gs:/u01

ON STANDBY SIDE:

• modify init parameter by adding these lines:

DB_NAME=ORCL

DB_UNIQUE_NAME=ORCL_STBY

FAL_SERVER=ORCL

FAL_CLIENT=ORCL_STBY

LOG_ARCHIVE_FORMAT='%t_%s_%r.arc'

STANDBY_FILE_MANAGEMENT=AUTO

LOG_ARCHIVE_CONFIG='dg_config=(ORCL,ORCL_STBY)'

LOG_ARCHIVE_DEST_1='LOCATION=/u01/db_files/ARCH'

Log_archive_dest_state_1=ENABLE

• add tnsnames.ora entries:

ORCL =

  (DESCRIPTION =

    (ADDRESS = (PROTOCOL = TCP)(HOST = oracle10g)(PORT = 1522))

    (CONNECT_DATA =

      (SERVER = DEDICATED)

      (SERVICE_NAME = ORCL)

    )

  )

CATALOG =

  (DESCRIPTION =

    (ADDRESS = (PROTOCOL = TCP)(HOST = oracle10g)(PORT = 1522))

    (CONNECT_DATA =

      (SERVER = DEDICATED)

      (SERVICE_NAME = CATALOG)

    )

  )

• start the standby listener

lsnrctl start

• start the instance in nomount state:

set environments:

export ORACLE_HOME=/u01/app/oracle/product/10.2.0/realdb_1

export ORACLE_SID=ORCL

export PATH=$PATH:$ORACLE_HOME/bin

$ sqlplus / as sysdba

startup nomount pfile=?/dbs/initORCL.ora

• restore database

$ rman catalog rman/rman@CATALOG target sys/******@ORCL auxiliary /

 run

{

allocate  channel C1 device type disk;

allocate  channel C2 device type disk;

set until SCN 1391907;

RESTORE DATABASE;

RECOVER DATABASE;

}

• start apply process: 

ALTER DATABASE RECOVER MANAGED STANDBY DATABASE DISCONNECT FROM SESSION;

If you need to cancel the process:

ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL;

If you need to open the standby process read only:

alter database open read only;

• Control scripts

Shows Data Guard log numbers:

Primary side:

select thread#, sequence# from v$log

Standby side

select thread#, sequence#, status from v$managed_standby;

Last archived log on standby

select max(sequence#) "Last Seq Recieved" from v$archived_log;

Shows archive processes status. If log_archive_dest_state_2 is VALID, it writes archive logs to standby side .

select dest_name,status,target,destination,db_unique_name from v$archive_dest;

Stop/start writing to standby side:

alter system set log_archive_dest_state_2=ENABLE

alter system set log_archive_dest_state_2=DEFER 

Dataguard messages:

select message from v$dataguard_status where dest_id = 2;