如何Shrink Undo表空间,释放过度占用的空间

环境:
OS:Red Hat Enterprise Linux AS release 4 (Nahant)
DB:Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Production
一台Oracle10gR2数据库报出如下错误:
ORA-1653: unable to extend table SYSMAN.MGMT_SYSTEM_ERROR_LOG by 8 in  tablespace SYSAUX
ORA-1653: unable to extend table SYSMAN.MGMT_SYSTEM_ERROR_LOG by 8 in  tablespace SYSAUX
ORA-1653: unable to extend table SYSMAN.MGMT_SYSTEM_ERROR_LOG by 8 in  tablespace SYSAUX
ORA-1653: unable to extend table SYSMAN.MGMT_SYSTEM_ERROR_LOG by 8 in  tablespace SYSAUX
ORA-1653: unable to extend table SYSMAN.MGMT_SYSTEM_ERROR_LOG by 8 in  tablespace SYSAUX

登陆检查，发现是SYSAUX表空间空间用尽,不能扩展,尝试手工扩展SYSAUX表空间:
 alter database datafile '+ORADG/danaly/datafile/sysaux.266.600173881' resize 800m
Tue Nov 29 23:31:38 2005
ORA-1237 signalled during: alter database datafile '+ORADG/danaly/datafile/sysaux.266.600173881' resize 800m...

出现ORA-1237错误，提示空间不足。这时候我才认识到是磁盘空间可能被用完了.

是谁"偷偷的"用了那么多空间呢(本来有几十个G的Free磁盘空间的)?
检查数据库表空间占用空间情况:
SQL> select tablespace_name,sum(bytes)/1024/1024/1024 GB
  2  from dba_data_files group by tablespace_name
  3  union all
  4  select tablespace_name,sum(bytes)/1024/1024/1024 GB
  5  from dba_temp_files group by tablespace_name order by GB;

TABLESPACE_NAME                        GB
------------------------------ ----------
USERS                          .004882813
UNDOTBS2                        .09765625
SYSTEM                         .478515625
SYSAUX                         .634765625
WAPCM_TS_VISIT_DETAIL            .9765625
HY_DS_DEFAULT                           1
MINT_TS_DEFAULT                         1
MMS_TS_DATA2                        1.375
MMS_IDX_SJH                             2
MMS_TS_DEFAULT                          2
IVRCN_TS_DATA                           2

TABLESPACE_NAME                        GB
------------------------------ ----------
MMS_TS_DATA1                            2
CM_TS_DEFAULT                           5
TEMP                           20.5498047
UNDOTBS1                       27.1582031

15 rows selected.
不幸的发现，UNDO表空间已经扩展至27G，而TEMP表空间也扩展至20G，这2个表空间加起来占用了47G的磁盘空间,导致了空间不足。
显然曾经有大事务占用了大量的UNDO表空间和Temp表空间，Oracle的AUM(Auto Undo Management)从出生以来就经常出现只扩展,不收缩(shrink)的情况(通常我们可以设置足够的UNDO表空间大小,然后取消其自动扩展属性).
现在我们可以采用如下步骤回收UNDO空间:

1.确认文件
SQL> select file_name,bytes/1024/1024 from dba_data_files
  2  where tablespace_name like 'UNDOTBS1';

FILE_NAME
--------------------------------------------------------------------------------
BYTES/1024/1024
---------------
+ORADG/danaly/datafile/undotbs1.265.600173875
          27810
2.检查UNDO Segment状态
SQL> select usn,xacts,rssize/1024/1024/1024,hwmsize/1024/1024/1024,shrinks
  2  from v$rollstat order by rssize;

       USN      XACTS RSSIZE/1024/1024/1024 HWMSIZE/1024/1024/1024    SHRINKS
---------- ---------- --------------------- ---------------------- ----------
         0          0            .000358582             .000358582          0
         2          0            .071517944             .071517944          0
         3          0             .13722229              .13722229          0
         9          0            .236984253             .236984253          0
        10          0            .625144958             .625144958          0
         5          1            1.22946167             1.22946167          0
         8          0            1.27175903             1.27175903          0
         4          1            1.27895355             1.27895355          0
         7          0            1.56770325             1.56770325          0
         1          0            2.02474976             2.02474976          0
         6          0             2.9671936              2.9671936          0

11 rows selected.
3.创建新的UNDO表空间
SQL> create undo tablespace undotbs2 ;
（经测试，在9i环境下后面还要加上datafile '/opt/..../undotbs2.dbf' size 1024M）
Tablespace created.
4.切换UNDO表空间为新的UNDO表空间
SQL> alter system set undo_tablespace=undotbs2 scope=both;

System altered.

创建了新的UNDO表空间以后，如果不知道系统使用的是pfile还是spfile文件，应使用参数both,会同时修改spfile文件，避免出现冲突。

5.等待原UNDO表空间所有UNDO SEGMENT OFFLINE
SQL> select usn,xacts,status,rssize/1024/1024/1024,hwmsize/1024/1024/1024,shrinks
  2 from v$rollstat order by rssize;


       USN      XACTS STATUS          RSSIZE/1024/1024/1024 HWMSIZE/1024/1024/1024    SHRINKS
---------- ---------- --------------- --------------------- ---------------------- ----------
        14          0 ONLINE                     .000114441             .000114441          0
        19          0 ONLINE                     .000114441             .000114441          0
        11          0 ONLINE                     .000114441             .000114441          0
        12          0 ONLINE                     .000114441             .000114441          0
        13          0 ONLINE                     .000114441             .000114441          0
        20          0 ONLINE                     .000114441             .000114441          0
        15          1 ONLINE                     .000114441             .000114441          0
        16          0 ONLINE                     .000114441             .000114441          0
        17          0 ONLINE                     .000114441             .000114441          0
        18          0 ONLINE                     .000114441             .000114441          0
         0          0 ONLINE                     .000358582             .000358582          0

       USN      XACTS STATUS          RSSIZE/1024/1024/1024 HWMSIZE/1024/1024/1024    SHRINKS
---------- ---------- --------------- --------------------- ---------------------- ----------
         6          0 PENDING OFFLINE             2.9671936              2.9671936          0

12 rows selected.
再看:
11:32:11 SQL> /

       USN      XACTS STATUS          RSSIZE/1024/1024/1024 HWMSIZE/1024/1024/1024    SHRINKS
---------- ---------- --------------- --------------------- ---------------------- ----------
        15          1 ONLINE                     .000114441             .000114441          0
        11          0 ONLINE                     .000114441             .000114441          0
        12          0 ONLINE                     .000114441             .000114441          0
        13          0 ONLINE                     .000114441             .000114441          0
        14          0 ONLINE                     .000114441             .000114441          0
        20          0 ONLINE                     .000114441             .000114441          0
        16          0 ONLINE                     .000114441             .000114441          0
        17          0 ONLINE                     .000114441             .000114441          0
        18          0 ONLINE                     .000114441             .000114441          0
        19          0 ONLINE                     .000114441             .000114441          0
         0          0 ONLINE                     .000358582             .000358582          0

11 rows selected.

Elapsed: 00:00:00.00
6.删除原UNDO表空间

11:34:00 SQL> drop tablespace undotbs1 including contents;
（本人经测试觉得其实是否可以用drop tablespace undotbs1 including contents and datafiles;直接连硬盘里面的dbf文件件一起删除）
Tablespace dropped.

Elapsed: 00:00:03.13
7.检查空间情况
由于我使用的ASM管理,可以使用10gR2提供的信工具asmcmd来察看空间占用情况.
[oracle@danaly ~]$ export ORACLE_SID=+ASM
[oracle@danaly ~]$ asmcmd
ASMCMD> du
Used_MB      Mirror_used_MB
  21625               21625
ASMCMD> exit

空间已经释放。