通过物理DG将单实例迁移至RAC

目的：将单实例的数据库（filesystem）迁移至RAC环境，停机时间非常有限，所以想到了通过物理DG来实现

环境准备：  RAC环境不需要安装数据库
单实例（归档模式）： 
IP： 192.168.10.26    
SID：ora11g
db_uniquen_name:ora11g

RAC:   
主机名：rac1,rac2
IP: 192.168.10.149 rac1
192.168.10.150 rac2

192.168.10.160 rac1-vip
192.168.10.161 rac2-vip

10.10.1.1 rac1-priv
10.10.1.2 rac2-priv

192.168.10.162 rac-scan

SID:  ora11g1  ora11g2
DB_UNIQUE_NAME:ora11g1

一、主库环境配置
1，修改主库为强制日志模式

点击(此处)折叠或打开

1. alter database force logging;

2，为主库添加standby日志（因为后面要switchover）

点击(此处)折叠或打开

1. alter database add standby logfile group 4 ('/oracle/app/oradata/ORA11G/standby_redo04.log') size 200m;
   
2. alter database add standby logfile group 5 ('/oracle/app/oradata/ORA11G/standby_redo05.log') size 200m;
   
3. alter database add standby logfile group 6 ('/oracle/app/oradata/ORA11G/standby_redo06.log') size 200m;
   
4. alter database add standby logfile group 7 ('/oracle/app/oradata/ORA11G/standby_redo07.log') size 200m;

3，复制密码文件到备库$ORACLE_HOME/dbs目录下

点击(此处)折叠或打开

1. scp $ORACLE_HOME/dbs/orapwora11g.ora rac1:$ORACLE_HOME/dbs

4，配置主库的tnsnames.ora，并复制至备库

点击(此处)折叠或打开

1. primary=
   
2.   (DESCRIPTION =
   
3.     (ADDRESS_LIST =
   
4.       (ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.10.26)(PORT = 1521))
   
5.     )
   
6.     (CONNECT_DATA =
   
7.       (SERVICE_NAME = ora11g)
   
8.     )
   
9.   )
   
10.  
    
11. standby=
    
12.   (DESCRIPTION =
    
13.     (ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.10.160)(PORT = 1521))
    
14.     (CONNECT_DATA =
    
15.       (SERVER = DEDICATED)
    
16.       (SERVICE_NAME = ora11g1)
    
17.       (UR = A)
    
18.     )
    
19.   )

注意：standby端即rac端的ip地址写的是ora11g1的vip地址，而且需要加上(UR=A)选项，否则通过sqlplus连接不到rac实例。

5，配置DG参数

点击(此处)折叠或打开

1. alter system set log_archive_config='dg_config=(ora11g,ora11g1)' scope=both;
   
2. alter system set log_archive_dest_1='location=/oracle/app/oracle/archive_log valid_for=(all_logfiles,all_roles) db_unique_name=ora11g' scope=both;
   
3. alter system set log_archive_dest_2='service=standby lgwr async valid_for=(online_logfiles,all_roles) db_unique_name=ora11g1' scope=both;
   
4. alter system set fal_server=ora11g1 scope=both;

6，生成pfile，并复制到备库

点击(此处)折叠或打开

1. SQL>create pfile from spfile;

点击( 此处)折叠或打开

1. scp $ORACLE_HOME/dbs/initora11g.ora rac1:$ORACLE_HOME/dbs

二、备库
1，修改主库复制过来的参数文件

点击(此处)折叠或打开

1. ora11g.__db_cache_size=335544320
   
2. ora11g.__java_pool_size=4194304
   
3. ora11g.__large_pool_size=4194304
   
4. ora11g.__oracle_base='/u01/app'
   
5. ora11g.__pga_aggregate_target=335544320
   
6. ora11g.__sga_target=503316480
   
7. ora11g.__shared_io_pool_size=0
   
8. ora11g.__shared_pool_size=146800640
   
9. ora11g.__streams_pool_size=0
   
10. *.audit_file_dest='/u01/app/oracle/admin/ora11g'
    
11. *.audit_trail='none'
    
12. *.compatible='11.2.0.1.0'
    
13. *.control_files='+DATA/ora11g/controlfile/control.ctl'
    
14. *.db_block_size=8192
    
15. *.db_domain=''
    
16. *.db_name='ora11g'
    
17. *.db_recovery_file_dest='+DATA'
    
18. *.db_recovery_file_dest_size=5218762752
    
19. *.diagnostic_dest='/u01/app/oracle'
    
20. *.memory_target=838860800
    
21. *.open_cursors=300
    
22. *.processes=1000
    
23. *.remote_login_passwordfile='EXCLUSIVE'
    
24. 
    
25. #### rac parameter
    
26. ora11g1.local_listener='(ADDRESS=(PROTOCOL=TCP)(HOST=192.168.10.160)(PORT=1521))'
    
27. ora11g2.local_listener='(ADDRESS=(PROTOCOL=TCP)(HOST=192.168.10.161)(PORT=1521))'
    
28. ora11g1.undo_tablespace='UNDOTBS1'
    
29. ora11g2.undo_tablespace='UNDOTBS2'
    
30. ora11g1.remote_listener='rac-scan:1521'
    
31. ora11g2.remote_listener='rac-scan:1521'
    
32. ora11g1.instance_name=ora11g1
    
33. ora11g2.instance_name=ora11g2
    
34. ora11g1.instance_number=1
    
35. ora11g2.instance_number=2
    
36. *.cluster_database=true
    
37. *.cluster_database_instances=2
    
38. *.db_create_file_dest='+DATA'
    
39. *.db_create_online_log_dest_1='+DATA'
    
40. ora11g1.thread=1
    
41. ora11g2.thread=2
    
42. 
    
43. ### dg parameter
    
44. *.db_unique_name=ora11g1
    
45. *.log_archive_config='dg_config=(ora11g1,ora11g)'
    
46. *.standby_file_management='AUTO'
    
47. *.db_file_name_convert='/oracle/app/oradata/ORA11G','+DATA/ora11g/datafile'
    
48. *.log_file_name_convert='/oracle/app/oradata/ORA11G','+DATA/ora11g/onlinelog'
    
49. *.log_archive_dest_1='location=+DATA/ora11g/archivelog valid_for=(all_logfiles,all_roles) db_unique_name=ora11g1'
    
50. *.log_archive_dest_2='service=primary lgwr async valid_for=(online_logfiles,all_roles) db_unique_name=ora11g'
    
51. *.fal_server=primary

2，备库创建需要的目录，并启动备库到nomount状态
文件系统目录：

点击(此处)折叠或打开

1. mkdir -p /u01/app/oracle/admin/ora11g/adump
2. mkdir -p /u01/app/oracle/admin/ora11g/dpdump
3. mkdir -p /u01/app/oracle/admin/ora11g/pfile

ASM目录：

点击(此处)折叠或打开

1. mkdir +DATA/ora11g/datafile
2. mkdir +DATA/ora11g/controlfile
3. mkdir +DATA/ora11g/onlinelog
4. mkdir +DATA/ora11g/archivelog

点击(此处)折叠或打开

1. SQL>startup nomount pfile='/u01/app/oracle/product/11.2.0/dbhome_1/dbs/initora11g1.ora'

3，测试tnsnames连通性
主库：

点击(此处)折叠或打开

1. sqlplus sys/*****@standby as sysdba

备库：

点击(此处)折叠或打开

1. sqlplus sys/*****@primary as sysdba

三、主库用rman备份

点击(此处)折叠或打开

1. rman target /
   
2. 
   
3. run {
   
4. allocate channel c1 device type disk;
   
5. allocate channel c2 device type disk;
   
6. backup incremental level 0
   
7. format '/home/oracle/rman_bak/inr0_%U' tag 'full_bak_for_rac_standby'
   
8. database plus archivelog delete all input;
   
9. release channel c1;
   
10. release channel c2;
    
11. }
12. 
    
13. backup format '/home/oracle/rman_bak/control01.ctl' current controlfile for standby;

四、通过duplicate复制至备库
rman target sys/*****@ora11g auxiliary sys/xj1234@standby
duplicate target database for standby from active database;
注意：1，11.2.0.1有个bug，用duplicate复制的时候，本机rman target / 这种方式会报错，需要写全用户名密码。
         2，备库的standby_file_management需要该问manaul模式，auto模式下日志rename会失败

五、备库启动redo应用：

点击(此处)折叠或打开

1. alter database recover managed standby database disconnect from session;
   
2. alter database recover managed standby database cancel;
   
3. alter database open read only;
   
4. alter database recover managed standby database using current logfile disconnect from session;

六、switch over，将备库转换为主库，分别在主库和备库进行角色转换：
主库：

点击(此处)折叠或打开

1. SELECT SWITCHOVER_STATUS FROM V$DATABASE;
   
2. ALTER DATABASE COMMIT TO SWITCHOVER TO PHYSICAL STANDBY WITH SESSION SHUTDOWN;

备库：

点击(此处)折叠或打开

1. SELECT SWITCHOVER_STATUS FROM V$DATABASE;
   
2. ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY with session shutdown;
   
3. select status from v$instance;
   
4. alter database open;

七、rac的一个节点已经可以正常打开数据库了，但是节点2需要的redo，undo表空间都还没有，需要手工创建：

点击(此处)折叠或打开

1. alter database add logfile thread 2 group 8 '+DATADG' size 100m;
   
2. alter database add logfile thread 2 group 9 '+DATADG' size 100m;
   
3. alter database add logfile thread 2 group 10 '+DATADG' size 100m;
   
4. alter database enable public thread 2;
   
5. @$ORACLE_HOME/rdbms/admin/catclust.sql

八、节点2打开数据库
SQL>startup  

九、此时的数据库是可以正常打开使用了，一般还需将参数文件写到asm中，而且此时的数据库都是本地管理，并没有注册到grid中，还需要手工注册：

点击(此处)折叠或打开

1. srvctl add database -d ora11g -o /u01/app/oracle/product/11.2.0/dbhome_1
2. srvctl config database
4. srvctl add instance -d ora11g -n rac1 -i ora11g1
5. srvctl add instance -d ora11g -n rac2 -i ora11g2

点击( 此处)折叠或打开

1. create spfile='+DATA' from pfile;

点击( 此处)折叠或打开

1. srvctl modify database -d ora11g -p '+data/ora11g1/parameterfile/spfile.325.886161615'

十、到此，单实例已经转换完成了rac数据库。可以重启一下机器，看看数据库能否自动启动，资源是否正常。
如有需要，可以修改service_name

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/29098758/viewspace-2139394/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/29098758/viewspace-2139394/