Oracle用户管理的不完全恢复：根据时间进行恢复

需要进行不完全恢复的场景：

　　a.因归档日志丢失而导致完全恢复失败。

　　b.所有未归档的重做日志文件和数据文件均丢失。

　　c.用户错误

　　　　1.某个重要的表被删除。　　2.在表中提交了无效数据。

　　d.当前控制文件丢失，必须使用备份控制文件才能打开数据库。

模拟场景：

1.1干净关闭数据库

SQL> conn /as sysdba
Connected.
SQL> shutdown immediate
Database closed.
Database dismounted.
ORACLE instance shut down.
SQL>

1.2 对数据库冷备

bash-3.00$ ls -l
total 2215712
-rw-r-----   1 oracle   oinstall 7389184 Jan 25 21:18 control01.ctl
-rw-r-----   1 oracle   oinstall 7389184 Jan 25 21:18 control02.ctl
-rw-r-----   1 oracle   oinstall 7389184 Jan 25 21:18 control03.ctl
-rw-r-----   1 oracle   oinstall 104865792 Jan 25 21:18 example01.dbf
-rw-r-----   1 oracle   oinstall 52429312 Jan 25 20:31 redo01.log
-rw-r-----   1 oracle   oinstall 52429312 Jan 25 21:18 redo02.log
-rw-r-----   1 oracle   oinstall 52429312 Jan 25 20:31 redo03.log
-rw-r-----   1 oracle   oinstall 10493952 Jan 25 21:18 sun01_1.dbf
-rw-r-----   1 oracle   oinstall 10493952 Jan 25 21:18 sun02_1.dbf
-rw-r-----   1 oracle   oinstall 5251072 Jan 25 21:18 sun03_1.dbf
-rw-r-----   1 oracle   oinstall 262152192 Jan 25 21:18 sysaux01.dbf
-rw-r-----   1 oracle   oinstall 503324672 Jan 25 21:18 system01.dbf
-rw-r-----   1 oracle   oinstall 20979712 Jan 23 04:03 temp01.dbf
-rw-r-----   1 oracle   oinstall 31465472 Jan 25 21:18 undotbs01.dbf
-rw-r-----   1 oracle   oinstall 5251072 Jan 25 21:18 users01.dbf
bash-3.00$ cp * /u01/backup/cold
bash-3.00$

1.3 在t表中插入数据

SQL> conn user1/user1
Connected.

SQL> select * from t;

        ID NAME
---------- ----------------
         1 oracle
         2 oracle
         0 oracle
         3 oracle
         4 oracle
         5 oracle

6 rows selected.

SQL> insert into t values(6,'oracle');

1 row created.

SQL> commit;

Commit complete.

SQL> 

1.4查看当前时间

SQL> select to_char(sysdate,'yyyy-mm-dd hh24:mi:ss') from dual;

TO_CHAR(SYSDATE,'YY
-------------------
2013-01-25 21:30:35

SQL> 

1.5 模拟失误现场

SQL> drop table t;

Table dropped.

SQL> 

恢复
2.1 关机

SQL> shutdown abort
ORACLE instance shut down.
SQL> 

2.2还原以前备份的数据文件

bash-3.00$ pwd
/u01/oradata/sunbak
bash-3.00$ ls -l
total 2215712
-rw-r-----   1 oracle   oinstall 7389184 Jan 25 21:33 control01.ctl
-rw-r-----   1 oracle   oinstall 7389184 Jan 25 21:33 control02.ctl
-rw-r-----   1 oracle   oinstall 7389184 Jan 25 21:33 control03.ctl
-rw-r-----   1 oracle   oinstall 104865792 Jan 25 21:21 example01.dbf
-rw-r-----   1 oracle   oinstall 52429312 Jan 25 21:21 redo01.log
-rw-r-----   1 oracle   oinstall 52429312 Jan 25 21:33 redo02.log
-rw-r-----   1 oracle   oinstall 52429312 Jan 25 21:21 redo03.log
-rw-r-----   1 oracle   oinstall 10493952 Jan 25 21:30 sun01_1.dbf
-rw-r-----   1 oracle   oinstall 10493952 Jan 25 21:21 sun02_1.dbf
-rw-r-----   1 oracle   oinstall 5251072 Jan 25 21:21 sun03_1.dbf
-rw-r-----   1 oracle   oinstall 262152192 Jan 25 21:33 sysaux01.dbf
-rw-r-----   1 oracle   oinstall 503324672 Jan 25 21:33 system01.dbf
-rw-r-----   1 oracle   oinstall 20979712 Jan 23 04:03 temp01.dbf
-rw-r-----   1 oracle   oinstall 31465472 Jan 25 21:33 undotbs01.dbf
-rw-r-----   1 oracle   oinstall 5251072 Jan 25 21:21 users01.dbf
bash-3.00$ rm -f *.dbf
bash-3.00$ cd /u01/backup/cold/
bash-3.00$ cp *.dbf  /u01/oradata/sunbak/
bash-3.00$ 

2.3将数据库开到mount

SQL> conn /as sysdba
Connected to an idle instance.
SQL> startup mount
ORACLE instance started.

Total System Global Area  289406976 bytes
Fixed Size                  1279820 bytes
Variable Size              92276916 bytes
Database Buffers          192937984 bytes
Redo Buffers                2912256 bytes
Database mounted.
SQL> 

2.4对比控制文件与数据文件的SCN

#最新的控制文件
SQL> select file#,checkpoint_change# from v$datafile;

     FILE# CHECKPOINT_CHANGE#
---------- ------------------
         1             690924
         2             690924
         3             690924
         4             690924
         5             690924
         6             690924
         7             690924
         8             690924

8 rows selected.

SQL> 
#备份的数据文件
SQL> select file#,checkpoint_change# from v$datafile_header;

     FILE# CHECKPOINT_CHANGE#
---------- ------------------
         1             690923
         2             690923
         3             690923
         4             690923
         5             690923
         6             690923
         7             690923
         8             690923

8 rows selected.

SQL> 

2.5 恢复

SQL> recover database until time '2013-01-25:21:30:35';  
Media recovery complete.
SQL> 
SQL> alter database open resetlogs;

Database altered.

SQL>

3.查看t表

SQL> select * from user1.t order by 1;

        ID NAME
---------- ----------------
         0 oracle
         1 oracle
         2 oracle
         3 oracle
         4 oracle
         5 oracle
         6 oracle

7 rows selected.

SQL>