Oracle UNDO Recreate & Recovery

Oracle UNDO表空间重建与恢复

UNDO表空间的数据文件从OS直接删除了，数据库还没shutduwn，UNDO需要恢复。

1，数据库正常运行状态中；
2，rm - fundotbs02.dbf， 数据库在不知情的情况下接着运行，直到有日志错误告警。

Fri Apr 19 13:27:03 2013[3942] ORA-01554 reason: failed to get txn-enqueue TX-0x000d0000-0x00000000Errors in file /u01/app/oracle/diag/rdbms/ora11g/ORA11G/trace/ORA11G_smon_3942.trc:ORA-01595: error freeing extent (17) of rollback segment (13))ORA-01554: transaction concurrency limit reached reason:failed to get TX-enqueue params:851968, 0Fri Apr 19 13:27:51 2013Errors in file /u01/app/oracle/diag/rdbms/ora11g/ORA11G/trace/ORA11G_j000_10253.trc:ORA-12012: error on auto execute of job 43ORA-01116: error in opening database file 7ORA-01110: data file 7: '/u01/app/oracle/ORA11G/undotbs02.dbf'ORA-27041: unable to open fileLinux Error: 2: No such file or directoryAdditional information: 3Errors in file /u01/app/oracle/diag/rdbms/ora11g/ORA11G/trace/ORA11G_j000_10253.trc:ORA-12012: error on auto execute of job 43ORA-01116: error in opening database file 7ORA-01110: data file 7: '/u01/app/oracle/ORA11G/undotbs02.dbf'ORA-27041: unable to open fileLinux Error: 2: No such file or directoryAdditional information: 3

3，此时创建spfile也会报错。
SQL> create pfile from spfile;
create pfile from spfile
*
ERROR at line 1:
ORA-01116: error in opening database file 7
ORA-01110: data file 7: '/u01/app/oracle/ORA11G/undotbs02.dbf'
ORA-27041: unable to open file
Linux Error: 2: No such file or directory
Additional information: 3
4，此时，注意保留现场，不要随意关闭数据库，查询有哪些UNDO段正在使用。
SQL> select * from v$rollname;
       USN NAME
---------- ------------------------------
         0 SYSTEM
        11 _SYSSMU11_2087506584$
        12 _SYSSMU12_1903386335$
        13 _SYSSMU13_3816552520$
        14 _SYSSMU14_2728152801$
        15 _SYSSMU15_2082928664$
        16 _SYSSMU16_1938921541$
        17 _SYSSMU17_3062294638$

8 rows selected.
如果数据库已经强行关闭了，使用隐含参数再次打开数据库后，可以进行如下查询，获取UNDO段信息。
SQL> select segment_name,status,tablespace_name from dba_rollback_segs where status='NEEDS RECOVERY';
   SEGMENT_NAME                   STATUS           TABLESPACE_NAME
------------------------------ ---------------- ------------------------------

        11 _SYSSMU11_2087506584$ NEEDS RECOVERY   UNDOTBS2
        12 _SYSSMU12_1903386335$ NEEDS RECOVERY   UNDOTBS2
        13 _SYSSMU13_3816552520$ NEEDS RECOVERY   UNDOTBS2
        14 _SYSSMU14_2728152801$ NEEDS RECOVERY   UNDOTBS2
        15 _SYSSMU15_2082928664$ NEEDS RECOVERY   UNDOTBS2
        16 _SYSSMU16_1938921541$ NEEDS RECOVERY   UNDOTBS2
        17 _SYSSMU17_3062294638$ NEEDS RECOVERY   UNDOTBS2

5，此时正常关闭数据库已经不行，只能强行关闭数据库。
SQL> shutdown immediate;
ORA-01116: error in opening database file 7
ORA-01110: data file 7: '/u01/app/oracle/ORA11G/undotbs02.dbf'
ORA-27041: unable to open file
Linux Error: 2: No such file or directory
Additional information: 3
shutdown abort
注意：
如果不能正常关闭数据库，意味着数据库状态出现了不一致，如果再次启动，此时需要启用下面的参数_offline_rollback_segments去绕过一致性检查。
如果能正常关闭数据库，意味着数据库状态一致，此时_offline_rollback_segments这个参数就不需要启用。

6，使用spfile的内容创建pfile，并对以下参数进行修改。

#*.undo_tablespace='UNDOTBS2'
undo_management='MANUAL'
rollback_segments='SYSTEM'
_offline_rollback_segments=(_SYSSMU11_2087506584$,_SYSSMU12_1903386335$,_SYSSMU13_3816552520$,_SYSSMU14_2728152801$,_SYSSMU15_2082928664$,_SYSSMU16_1938921541$,_SYSSMU17_3062294638$)

这里如果不使用_offline_rollback_segments参数，会出现下面的错误

SQL> alter database open;
alter database open
*
ERROR at line 1:
ORA-01157: cannot identify/lock data file 7 - see DBWR trace file
ORA-01110: data file 7: '/u01/app/oracle/ORA11G/undotbs02.dbf'
SQL> alter database datafile '/u01/app/oracle/ORA11G/undotbs02.dbf' offline drop;
Database altered.
SQL> alter database open;
Database altered.
SQL> drop tablespace UNDOTBS2;
drop tablespace UNDOTBS2
*
ERROR at line 1:
ORA-01548: active rollback segment '_SYSSMU11_2087506584$' found, terminate
dropping tablespace
7，从pfile启动数据库。

startup mount pfile='initORA11G.oa';

alter database datafile '/u01/app/oracle/ORA11G/undotbs02.dbf' offline drop;

alter database open;

drop tablespace UNDOTBS2;

8，重建UNDO表空间

create undo tablespace UNDOTBS3 datafile '/u01/app/oracle/ORA11G/undotbs03.dbf' size 1G;

shutdown immediate;

9，修改initORA11G.ora，将之前修改的参数改回来。

*.undo_tablespace='UNDOTBS3'

*.undo_management='AUTO'

10，启动数据

startup mount pfile='initORA11G.ora';

alter database open;

create spfile from pfile;

shutdown immediate;

startup;

UNDO表空间的数据文件从OS直接删除了，数据库已经shutduwn，UNDO需要恢复。

1，通过备份进行完全恢复。

2，通过上面的UNDO参数修改，重建UNDO表空间恢复。

无论哪种方式恢复，建议首先保留现场，进行冷备份。