ORACLE 12C DataGuard最大保护模式不需要终止主库

                   ORACLE 12C DataGuard最大保护模式不需要终止主库


 oracle 12c的DG处理maximize protection mode时,当备库异常关闭是,主库不会被关闭
在alert.log一直报异常需要人为手动干预


当前数据库保护模式
DONGDONGTANG> select protection_mode from v$database;


PROTECTION_MODE
--------------------
MAXIMUM PROTECTION




在standby上手动kill掉进程摸拟失败情况


[root@dg2 ~]# kill -9 22638


Sun Apr 03 05:58:52 2016
MMNL (ospid: 22676): terminating the instance due to error 472
Sun Apr 03 05:58:52 2016


在主库的alert.log中,只报连接不上从库


ARC2: Attempting destination LOG_ARCHIVE_DEST_2 network reconnect (3113)
Error 1034 received logging on to the standby
ARC2: Error 1034 attaching to RFS for reconnect
PING[ARC2]: Error 3113 when pinging standby dc1stby.
Sun Apr 03 09:37:14 2016
Error 1034 received logging on to the standby
Sun Apr 03 09:37:14 2016
LGWR: Error 1034 attaching to RFS for reconnect




如果是11g的版本,这个时候主库会直接被lgwr进程终止.




Destination LOG_ARCHIVE_DEST_2 is UNSYNCHRONIZED


LGWR: All standby destinations have failed


******************************************************
WARNING: All standby database destinations have failed
WARNING: Instance shutdown required to protect primary
******************************************************


LGWR (ospid: 25830): terminating the instance due to error 16098


Instance terminated by LGWR, pid = 25830




在12c中主库不会再强制结束了,等待人为去干预,这时候无法执行事务,只能做只读查询.
相比以前的版本来说这样更加合理了.一,不会是主库被强制结束,重启不会需要实例恢复;
二,有时候可能只是短暂的故障,比如网络异常.数据库自动重连,这样又会恢复正常; 
三,还可以执行只读查询.