控制文件发生损坏，如何恢复

1、You may need to create control files if:

All control files are lost because of a failure.

The name of a database needs to be changed.

The current settings in the control file need to be changed.

2、The behavior of multiplexed control files is this:

The database writes to all filenames listed for the initialization parameter CONTROL_FILES in the database initializaton parameter file.

oracle server 在往控制文件中写数据的时候，会同时往多个文件中写数据，而且内容是一样的.

The database reads only the first file listed in the CONTROL_FILES parameter during database operation.

oracle server 在读控制文件的时候，只会读第一个控制文件，其他的不读.

If any of the control files become unavailable during database operation,the instance becomes inoperable and should be aborted.

如果控制文件中的任何一个失效或者损坏，那么instance将会停止工作，此时DBA应当shutdown abort；

3、对比动态性能试图

SQL> select file#,checkpoint_change# from v$datafile_header;

FILE# CHECKPOINT_CHANGE#
---------- ------------------
1            1464420
2            1464420
3            1464420
4            1464420
5            1464420
6            1464420
7            1465263

7 rows selected.

动态性能试图v$datafile_header是通过读数据文件的头信息获取到的scn号

SQL> select file#,checkpoint_change# from v$datafile;

FILE# CHECKPOINT_CHANGE#
---------- ------------------
1            1464420
2            1464420
3            1464420
4            1464420
5            1464420
6            1464420
7            1465263

7 rows selected.

动态性能试图v$datafile是通过读控制文件的信息，获取到scn号的.

4、SCN Scenarios

Condition on Start-Up	Oracle Behavior	DBA Action
CF checkpoint SCN<Datafile checkpoint SCN	"Control file too old" error	Restore a newer control file.
CF checkpoint SCN>Datafile checkpoint SCN	Media recovery required	Most likely a datafile has been restored from a backup.Recovery is now required.
(CF checkpoint SCN=Datafile SCN) Database in mount mode,instance thread status=OPEN	Start up normally Crash recovery required	None None

5、控制文件恢复的方法

（a）、控制文件丢失一个，其他文件完好无损，直接拷贝另外的控制文件或者修改spfile文件都可以重新恢复数据库；

（b）、所有的控制文件都损坏，根据trace file进行脚本执行，然后恢复控制文件即可；