1.停止Redo Apply
如果備庫正處於Redo Apply過程，需要先取消。
sys@ora11gdg> alter database recover managed standby database cancel;
Database altered.
2.查看當前備庫狀態確保備庫處於MOUNTED狀態
sys@ora11gdg> select database_role,open_mode from v$database;
DATABASE_ROLE OPEN_MODE
---------------- --------------------
PHYSICAL STANDBY MOUNTED
此時備庫是物理備庫角色，運行模式是MOUNTED。
3.確保閃回恢復區已指定
友情提示：實現Snapshot Standby數據庫功能並不需要開啓主庫和備庫的閃回數據庫（Flashback Database）功能，與是否開啓閃回數據庫無關。
sys@ora11gdg> show parameter db_recovery_file_dest
NAME TYPE VALUE
--------------------------- ------------ ------------------------------------
db_recovery_file_dest string /u01/app/oracle/flash_recovery_area
db_recovery_file_dest_size big integer 3852M
確認主庫閃回功能並未開啓
sys@ora11g> select FLASHBACK_ON from v$database;
FLASHBACK_ON
------------------
NO
確認備庫閃回功能並未開啓
sys@ora11gdg> select FLASHBACK_ON from v$database;
FLASHBACK_ON
------------------
NO
4.調整備庫到Snapshot Standby數據庫狀態
只需要執行一條非常簡單的SQL命令便可以將備庫調整到Snapshot Standby數據庫。
sys@ora11gdg> alter database convert to snapshot standby;
Database altered.
sys@ora11gdg> select database_role,open_mode from v$database;
DATABASE_ROLE OPEN_MODE
---------------- --------------------
SNAPSHOT STANDBY MOUNTED
5.將備庫置於對外可讀寫狀態
sys@ora11gdg> alter database open;
Database altered.
sys@ora11gdg> select database_role,open_mode from v$database;
DATABASE_ROLE OPEN_MODE
---------------- --------------------
SNAPSHOT STANDBY READ WRITE
一套全新的可讀寫數據庫展現在我們面前。
6.分析切換過程中的日誌信息
ora11g主庫alert日誌：
Mon July 19 18:46:28 2013
LNS: Attempting destination LOG_ARCHIVE_DEST_2 network reconnect (3135)
LNS: Destination LOG_ARCHIVE_DEST_2 network reconnect abandoned
Errors in file /u01/app/oracle/diag/rdbms/ora11g/ora11g/trace/ora11g_nsa2_27302.trc:
ORA-03135: connection lost contact
Error 3135 for archive log file 2 to 'ora11gdg'
Errors in file /u01/app/oracle/diag/rdbms/ora11g/ora11g/trace/ora11g_nsa2_27302.trc:
ORA-03135: connection lost contact
LNS: Failed to archive log 2 thread 1 sequence 50 (3135)
Errors in file /u01/app/oracle/diag/rdbms/ora11g/ora11g/trace/ora11g_nsa2_27302.trc:
ORA-03135: connection lost contact
ora11gdg備庫alert日誌：
Mon July 19 18:46:26 2013
alter database convert to snapshot standby
Starting background process RVWR
Mon July 19 18:46:26 2013
RVWR started with pid=26, OS id=8824
Allocated 3981204 bytes in shared pool for flashback generation buffer
Created guaranteed restore point SNAPSHOT_STANDBY_REQUIRED_03/19/2013 18:46:26
krsv_proc_kill: Killing 3 processes (all RFS)
Begin: Standby Redo Logfile archival
End: Standby Redo Logfile archival
RESETLOGS after complete recovery through change 1472476
Resetting resetlogs activation ID 4174194338 (0xf8cd26a2)
Online log /u01/app/oracle/oradata/ora11gdg/redo01.log: Thread 1 Group 1 was previously cleared
Online log /u01/app/oracle/oradata/ora11gdg/redo02.log: Thread 1 Group 2 was previously cleared
Online log /u01/app/oracle/oradata/ora11gdg/redo03.log: Thread 1 Group 3 was previously cleared
Standby became priJulyy SCN: 1472474
Mon July 19 18:46:29 2013
Setting recovery target incarnation to 5
CONVERT TO SNAPSHOT STANDBY: Complete - Database mounted as snapshot standby
Completed: alter database convert to snapshot standby
關鍵的一行提示信息“Created guaranteed restore point SNAPSHOT_STANDBY_REQUIRED_03/19/2013 18:46:26”，這裏給出了我們轉換成snapshot的時刻，便於後面的回切。
7.測試備庫處於Snapshot Standby數據庫對主庫日誌的接收
當主庫切換日誌時，備庫依然可以接收到日誌，只是並不應用
1）主庫切換日誌
sys@ora11g> alter system switch logfile;
System altered.
2）主庫記錄的alert日誌內容
ora11g主庫alert日誌：
Mon July 19 18:52:00 2013
Thread 1 cannot allocate new log, sequence 52
Private strand flush not complete
Current log# 3 seq# 51 mem# 0: /u01/app/oracle/oradata/ora11g/redo03.log
Mon July 19 18:52:00 2013
ARC3: Standby redo logfile selected for thread 1 sequence 50 for destination LOG_ARCHIVE_DEST_2
Thread 1 advanced to log sequence 52 (LGWR switch)
Current log# 1 seq# 52 mem# 0: /u01/app/oracle/oradata/ora11g/redo01.log
Mon July 19 18:52:03 2013
Archived Log entry 91 added for thread 1 sequence 51 ID 0xf8cd26a2 dest 1:
Mon July 19 18:52:03 2013
LNS: Standby redo logfile selected for thread 1 sequence 51 for destination LOG_ARCHIVE_DEST_2
LNS: Standby redo logfile selected for thread 1 sequence 52 for destination LOG_ARCHIVE_DEST_2
ora11gdg備庫alert日誌：
Mon July 19 18:52:00 2013
RFS[5]: Assigned to RFS process 9174
RFS[5]: Identified database type as 'snapshot standby': Client is ARCH pid 27296
Mon July 19 18:52:00 2013
RFS[6]: Assigned to RFS process 9176
RFS[6]: Identified database type as 'snapshot standby': Client is ARCH pid 27300
RFS[6]: Selected log 4 for thread 1 sequence 50 dbid -120744030 branch 778023141
Mon July 19 18:52:00 2013
Archived Log entry 47 added for thread 1 sequence 50 ID 0xf8cd26a2 dest 1:
Mon July 19 18:52:03 2013
RFS[7]: Assigned to RFS process 9180
RFS[7]: Identified database type as 'snapshot standby': Client is LGWR ASYNC pid 27302
RFS[7]: Selected log 4 for thread 1 sequence 51 dbid -120744030 branch 778023141
Mon July 19 18:52:04 2013
Archived Log entry 48 added for thread 1 sequence 51 ID 0xf8cd26a2 dest 1:
RFS[7]: Selected log 4 for thread 1 sequence 52 dbid -120744030 branch 778023141
3）查看主庫和備庫歸檔目錄下的日誌文件內容
（1）主庫歸檔日誌文件
ora11g@secdb /home/oracle/arch/ora11g$ ls -ltr
total 879M
……省略其他……
-rw-r----- 1 oracle oinstall 1.1M July 19 18:51 1_50_778023141.arc
-rw-r----- 1 oracle oinstall 363K July 19 18:52 1_51_778023141.arc
（2）備庫歸檔日誌文件
ora11g@secdb /home/oracle/arch/ora11gdg$ ls -ltr
total 847M
……省略其他……
-rw-r----- 1 oracle oinstall 1.1M July 19 18:52 1_50_778023141.arc
-rw-r----- 1 oracle oinstall 363K July 19 18:52 1_51_778023141.arc
可見，備庫已經接受到主庫發過來的日誌。
8.在Snapshot Standby數據創建用戶和表並初始化數據
sys@ora11gdg> create user ocmu identified by ocmu;
User created.
secooler@ora11gdg> grant dba to ocmu;
Grant succeeded.
secooler@ora11gdg> conn ocmu/ocmu
Connected.
ocmu@ora11gdg> create table t (x varchar2(8));
Table created.
ocmu@ora11gdg> insert into t values ('Secooler');
1 row created.
ocmu@ora11gdg> commit;
Commit complete.
ocmu@ora11gdg> select * from t;
X
--------
Secooler
結論，此時備庫是一個可任意修改和調整的狀態，也就是我們要的“READ WRITE”可讀寫狀態。
特別注意的是，原理上實現Snapshot Standby數據庫功能是基於閃回數據原理的，因此任何導致閃回數據庫無法回退的動作在這裏也要規避，否則Snapshot Standby數據庫將無法回到曾經的備庫恢復狀態。
9.恢復Snapshot Standby數據庫爲Physical Standby數據庫
1）重啓備庫到MOUNTED狀態
ocmu@ora11gdg> conn / as sysdba
Connected.
sys@ora11gdg> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
sys@ora11gdg> startup mount
ORACLE instance started.
Total System Global Area 313860096 bytes
Fixed Size 1336232 bytes
Variable Size 268438616 bytes
Database Buffers 37748736 bytes
Redo Buffers 6336512 bytes
Database mounted.
sys@ora11gdg> select database_role,open_mode from v$database;
DATABASE_ROLE OPEN_MODE
---------------- --------------------
SNAPSHOT STANDBY MOUNTED
2）一條命令恢復原物理備庫身份
sys@ora11gdg> alter database convert to physical standby;
Database altered.
3）備庫的alert日誌清楚的記錄了這個切換的過程
Mon July 19 19:30:24 2013
alter database convert to physical standby
ALTER DATABASE CONVERT TO PHYSICAL STANDBY (ora11gdg)
Flashback Restore Start
Flashback Restore Complete
Stopping background process RVWR
Deleted Oracle managed file /u01/app/oracle/flash_recovery_area/ORA11GDG/flashback/o1_mf_7pg3n2jc_.flb
Deleted Oracle managed file /u01/app/oracle/flash_recovery_area/ORA11GDG/flashback/o1_mf_7pg52yst_.flb
Guaranteed restore point dropped
Clearing standby activation ID 4174523254 (0xf8d22b76)
The priJulyy database controlfile was created using the
'MAXLOGFILES 30' clause.
There is space for up to 27 standby redo logfiles
Use the following SQL commands on the standby database to create
standby redo logfiles that match the priJulyy database:
ALTER DATABASE ADD STANDBY LOGFILE 'srl1.f' SIZE 52428800;
ALTER DATABASE ADD STANDBY LOGFILE 'srl2.f' SIZE 52428800;
ALTER DATABASE ADD STANDBY LOGFILE 'srl3.f' SIZE 52428800;
ALTER DATABASE ADD STANDBY LOGFILE 'srl4.f' SIZE 52428800;
Completed: alter database convert to physical standby
從alert日誌中可以得到恢復方法使用的閃回數據庫功能實現的，也就是說，即便備庫沒有運行在閃回數據庫狀態，依然可以使用閃回數據庫功能完成備庫的角色轉換。
4）重啓備庫到自動恢復日誌狀態
（1）此時數據庫處於NOMOUNTED狀態，需要重新啓動數據庫。
注意這裏是重啓數據庫，而不是使用alter命令調整，否則會收到如下報錯：
sys@ora11gdg> alter database mount;
alter database mount
*
ERROR at line 1:
ORA-00750: database has been previously mounted and dismounted
sys@ora11gdg> shutdown immediate;
ORA-01507: database not mounted
ORACLE instance shut down.
sys@ora11gdg> startup mount;
ORACLE instance started.
Total System Global Area 313860096 bytes
Fixed Size 1336232 bytes
Variable Size 268438616 bytes
Database Buffers 37748736 bytes
Redo Buffers 6336512 bytes
Database mounted.
sys@ora11gdg> alter database recover managed standby database disconnect;
Database altered.
（2）查看備庫alert日誌，可以清楚的看到恢復的過程。
Mon July 19 19:43:48 2013
Managed Standby Recovery not using Real Time Apply
Parallel Media Recovery started with 4 slaves
Waiting for all non-current ORLs to be archived...
All non-current ORLs have been archived.
Clearing online redo logfile 1 /u01/app/oracle/oradata/ora11gdg/redo01.log
Clearing online log 1 of thread 1 sequence number 1
Completed: alter database recover managed standby database disconnect
Clearing online redo logfile 1 complete
Clearing online redo logfile 2 /u01/app/oracle/oradata/ora11gdg/redo02.log
Clearing online log 2 of thread 1 sequence number 2
Clearing online redo logfile 2 complete
Media Recovery Log /home/oracle/arch/ora11gdg/1_49_778023141.arc
Media Recovery Log /home/oracle/arch/ora11gdg/1_50_778023141.arc
Media Recovery Log /home/oracle/arch/ora11gdg/1_51_778023141.arc
Media Recovery Log /home/oracle/arch/ora11gdg/1_52_778023141.arc
Media Recovery Log /home/oracle/arch/ora11gdg/1_53_778023141.arc
Media Recovery Log /home/oracle/arch/ora11gdg/1_54_778023141.arc
Media Recovery Waiting for thread 1 sequence 55
（3）查看V$ARCHIVED_LOG動態性能視圖查看日誌應用情況
sys@ora11gdg> select sequence#, first_time, next_time, applied from v$archived_log order by sequence#;
SEQUENCE# FIRST_TIME NEXT_TIME APPLIED
---------- ----------------- ----------------- ---------
……省略其他數據……
49 20130720 18:32:32 20130720 18:38:03 YES
50 20130720 18:38:03 20130720 18:51:00 YES
51 20130720 18:51:00 20130720 18:52:03 YES
52 20130720 18:52:03 20130720 19:09:57 YES
53 20130720 19:09:57 20130720 19:10:15 YES
54 20130720 19:10:15 20130720 19:10:25 YES
52 rows selected.
10.開啓備庫到READ ONLY狀態驗證之前在Snapshot Standby數據庫上的操作已撤銷
sys@ora11gdg> alter database recover managed standby database cancel;
Database altered.
sys@ora11gdg> alter database open read only;
Database altered.
sys@ora11gdg> select database_role,open_mode from v$database;
DATABASE_ROLE OPEN_MODE
---------------- --------------------
PHYSICAL STANDBY READ ONLY
sys@ora11gdg> select username from dba_users where username = 'OCMU';
no rows selected
之前創建的測試用戶OCMU不存在。結論得證。
本文出自 “無雙城” 博客，請務必保留此出處http://929044991.blog.51cto.com/1758347/1255180