[20160222]Oracle 11G Data Guard Failover-flush redo.txt

--链接: http://blog.csdn.net/tianlesoftware/article/details/6256542

在Oracle 11g里，Data Guard 切换多了一个新的功能：flush redo。

Flush 能把没有发送的redo 从主库传送到standby库。 只要主库能启动到mount 状态，那么Flush 就可以把没有发送的归档和current
online redo 发送到备库。

Flush语法：

SQL> ALTER SYSTEM FLUSH REDO TO target_db_name;

这里的target_db_name 是我们在主库的db_unique_name 名称。 也就是在tnsnames.ora 文件配置的。 Flush 会将未发送的redo 从主库
传到备库，并且等待redo 在standby 库上apply 之后返回成功。 所以只要Flush成功，那么Failover 就没有主句丢失。

--自己测试看看：

1.主库：
SYS@test> show parameter unique
NAME            TYPE    VALUE
--------------- ------- -------
db_unique_name  string  test

--关闭数据库使用shutdown abort关闭。

SYS@test> shutdown abort ;
ORACLE instance shut down.

SYS@test> startup mount
ORACLE instance started.
Total System Global Area 1603411968 bytes
Fixed Size                  2228784 bytes
Variable Size             973082064 bytes
Database Buffers          620756992 bytes
Redo Buffers                7344128 bytes
Database mounted.

2.在备库执行：
SYS@testdg> ALTER SYSTEM FLUSH REDO TO 'test';
System altered.

--注意test要加引号，不然报错：
SYS@testdg> ALTER SYSTEM FLUSH REDO TO test;
ALTER SYSTEM FLUSH REDO TO test
                           *
ERROR at line 1:
ORA-00922: missing or invalid option

$ oerr ora 922
00922, 00000, "missing or invalid option"
// *Cause:
// *Action:

--检查alert*.log发现：

ALTER SYSTEM FLUSH REDO TO 'test' CONFIRM APPLY
Flush redo operation not allowed on a physical standby

--估计dg开在open read only模式，在active data guard状态。启动到mount状态继续测试：

SYS@testdg> shutdown immediate ;
Database closed.
Database dismounted.
ORACLE instance shut down.

SYS@testdg> startup mount
ORACLE instance started.
Total System Global Area 1603411968 bytes
Fixed Size                  2228784 bytes
Variable Size             905973200 bytes
Database Buffers          687865856 bytes
Redo Buffers                7344128 bytes
Database mounted.

SYS@testdg> ALTER SYSTEM FLUSH REDO TO 'test';
System altered.

--检查alert*.log发现：

Mon Feb 22 16:37:01 2016
ALTER SYSTEM FLUSH REDO TO 'test' CONFIRM APPLY
Flush redo operation not allowed on a physical standby
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Using STANDBY_ARCHIVE_DEST parameter default value as /u01/app/oracle11g/archivelog
ALTER SYSTEM SET log_archive_trace=0 SCOPE=BOTH SID='testdg';
ALTER SYSTEM SET log_archive_format='%t_%s_%r.dbf' SCOPE=SPFILE SID='testdg';
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE  THROUGH ALL SWITCHOVER DISCONNECT  USING CURRENT LOGFILE
Attempt to start background Managed Standby Recovery process (testdg)
Mon Feb 22 16:37:04 2016
MRP0 started with pid=29, OS id=1006
MRP0: Background Managed Standby Recovery process started (testdg)
started logmerger process
Mon Feb 22 16:37:09 2016
Managed Standby Recovery starting 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.
Recovery of Online Redo Log: Thread 1 Group 5 Seq 4815 Reading mem 0
  Mem# 0: /u01/app/oracle11g/oradata/test/redostb02.log
Completed: ALTER DATABASE RECOVER MANAGED STANDBY DATABASE  THROUGH ALL SWITCHOVER DISCONNECT  USING CURRENT LOGFILE
--注意看~,感觉还是不对。

3.感觉应该在主库执行：
SYS@test> ALTER SYSTEM FLUSH REDO TO 'testdg';
System altered.

--检查备库的alert*.log:
Mon Feb 22 16:43:48 2016
Standby switchover readiness check: Checking whether recoveryapplied all redo..
Physical Standby applied all the redo from the primary.

--如果备库打开read only：

SYS@test> ALTER SYSTEM FLUSH REDO TO 'testdg';
ALTER SYSTEM FLUSH REDO TO 'testdg'
*
ERROR at line 1:
ORA-16447: Redo apply was not active at the target standby database

SYS@test> ALTER SYSTEM FLUSH REDO TO testdg;
ALTER SYSTEM FLUSH REDO TO testdg
                           *
ERROR at line 1:
ORA-00922: missing or invalid option
--必须加引号。

--说明原链接的介绍存在问题。