本节将演示下rman基于时间点的不完全恢复，修改过的数据有执行commit操作的事务，只要备份文件和联机重做日志或者归档日志文件存在，理论上都是可以恢复的
1:使用rman备份test表空间，test在oracle里面是一个保留关键字，所以备份的时候要加双引号
RMAN> backup tablespace "TEST" format '/u01/backup/test_1216.dbf' tag ts;
2：使用yang用户连接数据库实例，进行数据操作，改用户的默认表空间即为test
yang SQL>select default_tablespace from dba_users where username='YANG';
DEFAULT_TABLESPACE
------------------------------
TEST
yang SQL>conn yang/yang
Connected.
yang SQL>col object_name format a30;
yang SQL>select object_name,object_type from user_objects;
OBJECT_NAME OBJECT_TYPE
------------------------------ -------------------
TEST TABLE
yang SQL>select * from test order by 1;
ID NAME
---------- --------------------------------
1 one
2 two
3 three
3:插入一条新数据，并提交事务
yang SQL>insert into test values (4,'four');
1 row created.
yang SQL>commit;
Commit complete.
4：使用sys用户登录数据库，查看当前的联机日志组为组1
yang SQL>show user;
USER is "SYS"
yang SQL>select group#,status from v$log;
GROUP# STATUS
---------- ----------------
1 CURRENT
2 UNUSED
3 UNUSED
yang SQL>!strings /u01/app/oracle/oradata/yang/redo01.log |grep 'four' //在lgwr进程发生之前，记录是保存在redolog buffer中
yang SQL>!strings /u01/app/oracle/oradata/yang/redo01.log |grep 'four' //commit后，记录保存在当前active状态的联机重做日志文件中
four
yang SQL>!strings /u01/app/oracle/oradata/yang/test01.dbf |grep 'four'
yang SQL>alter system checkpoint; //在checkpoint发生后，记录才真实可靠的写入到数据文件中
System altered.
yang SQL>!strings /u01/app/oracle/oradata/yang/test01.dbf |grep 'four'
four,
5：继续更新一条记录，但不执行提交操作
yang SQL>insert into test values (5,'five');
1 row created.
yang SQL>select count(*) from test;
COUNT(*)
----------
5
6：记录正常时刻的系统时间，后面用来做恢复操作时使用，生产环境下如果遇到类似的情况，可以根据实际情况进行估计，通常使用logminer工具获取这个时间
yang SQL>select to_char(sysdate,'YYYY-MM-DD:HH24:MI:SS') as time from dual;
TIME
-------------------
2010-12-16:15:37:05
7：模拟数据丢失，这里就直接删除test表，在生产环境下，可能出现磁盘故障，用户误操作等情况使关键的数据丢失
yang SQL>drop table test;
Table dropped.
yang SQL>select count(*) from test;
select count(*) from test
*
ERROR at line 1:
ORA-00942: table or view does not exist
8:将数据库关闭，重新启动到mount状态，并使用rman进行恢复
yang SQL>conn /as sysdba
Connected.
yang SQL>shutdown immediate;
Database closed.
Database dismounted.
ORACLE instance shut down.
[oracle@rhel6 oradata]$ rman target /
Recovery Manager: Release 11.2.0.1.0 - Production on Thu Dec 16 16:01:43 2010
Copyright (c) 1982, 2009, Oracle and/or its affiliates. All rights reserved.
connected to target database: YANG (DBID=3456695003, not open)
RMAN> startup mount;
Oracle instance started
database mounted
Total System Global Area 1068937216 bytes
Fixed Size 2220200 bytes
Variable Size 801115992 bytes
Database Buffers 260046848 bytes
Redo Buffers 5554176 bytes
RMAN> run {
2> allocate channel c1 type disk; //分配三个通道，提高并发
3> allocate channel c2 type disk;
4> allocate channel c3 type disk;
5> set until time "to_date('2010-12-16:15:37:05','YYYY-MM-DD:HH24:MI:SS')"; //指定恢复数据库状态的时间点
6> restore database; //还原数据库文件
7> recover database; //恢复数据库文件
8> alter database open resetlogs; //将数据库打到open状态，resetlogs参数将日志重新归位,使数据库的scn，检查点等保持一致
9> }
using target database control file instead of recovery catalog
allocated channel: c1
channel c1: SID=6 device type=DISK
allocated channel: c2
channel c2: SID=89 device type=DISK
allocated channel: c3
channel c3: SID=8 device type=DISK
executing command: SET until clause
Starting restore at 16-DEC-10
channel c1: starting datafile backup set restore
channel c1: specifying datafile(s) to restore from backup set
channel c1: restoring datafile 00001 to /u01/app/oracle/oradata/yang/system01.dbf
channel c1: restoring datafile 00002 to /u01/app/oracle/oradata/yang/sysaux01.dbf
channel c1: restoring datafile 00003 to /u01/app/oracle/oradata/yang/undotbs01.dbf
channel c1: restoring datafile 00004 to /u01/app/oracle/oradata/yang/users01.dbf
channel c1: restoring datafile 00006 to /u01/app/oracle/oradata/yang/backup_test1.dbf
channel c1: reading from backup piece /u01/backup/full_bak1216.dbf
channel c2: starting datafile backup set restore
channel c2: specifying datafile(s) to restore from backup set
channel c2: restoring datafile 00005 to /u01/app/oracle/oradata/yang/test01.dbf
channel c2: reading from backup piece /u01/backup/test_1216.dbf
channel c2: piece handle=/u01/backup/test_1216.dbf tag=TS
channel c2: restored backup piece 1
channel c2: restore complete, elapsed time: 00:00:03
channel c1: piece handle=/u01/backup/full_bak1216.dbf tag=FULL_BACKUP
channel c1: restored backup piece 1
channel c1: restore complete, elapsed time: 00:00:45
Finished restore at 16-DEC-10
Starting recover at 16-DEC-10
starting media recovery
archived log for thread 1 with sequence 1 is already on disk as file /u01/app/oracle/flash_recovery_area/YANG/archivelog/2010_12_16/o1_mf_1_1_6jmc341s_.arc
archived log file name=/u01/app/oracle/flash_recovery_area/YANG/archivelog/2010_12_16/o1_mf_1_1_6jmc341s_.arc thread=1 sequence=1
media recovery complete, elapsed time: 00:00:08
Finished recover at 16-DEC-10
database opened
released channel: c1
released channel: c2
released channel: c3
9：测试恢复数据
SQL>conn yang/yang
Connected.
yang SQL>desc test;
Name Null? Type
----------------------------------------- -------- ----------------------------
ID NUMBER(38)
NAME VARCHAR2(32)
yang SQL>select * from test;
ID NAME
---------- --------------------------------
3 three
1 one
2 two
4 four
