ITPub博客

首页 > Linux操作系统 > Linux操作系统 > 数据库UNDO损坏

数据库UNDO损坏

原创 Linux操作系统 作者:dear2170 时间:2009-03-23 09:51:50 0 删除 编辑

DB:ORACLE9206
OS:AIX5304
症状:数据库每3分钟产生一个归档,每个归档200M,数据库大小160G。应用(EBS)无法正常使用。手动停应用也停不下了,命令hang住不动。数据库shutdown immediate也停不下来,也是hang住不动,只能abort关掉(PRD环境,)查看数据库日志,发现全是错误:

SMON: about to recover undo segment 8
SMON: mark undo segment 8 as available
SMON: about to recover undo segment 8
SMON: mark undo segment 8 as available
SMON: about to recover undo segment 8
SMON: mark undo segment 8 as available
SMON: about to recover undo segment 8
SMON: mark undo segment 8 as available
SMON: about to recover undo segment 8
SMON: mark undo segment 8 as available
SMON: about to recover undo segment 8
SMON: mark undo segment 8 as available
SMON: about to recover undo segment 8
SMON: mark undo segment 8 as available
SMON: about to recover undo segment 8
SMON: mark undo segment 8 as available

并不停的刷新

判断是undo表空间出问题,决定重建undo表空间,步骤如下:

1:将数据库abort掉之后startup起来。

2:删掉原undo表空间

SQL> drop tablespace APPS_UNDOTS1;

Tablespace dropped.

3:将pfile里的如下undo信息注释掉

#undo_management=AUTO

#undo_tablespace=APPS_UNDOTS1

4:重启数据库,还是需要abort掉后起来。

5:创建新的undo表空间

SQL> create undo tablespace APPS_UNDOTS2  datafile '/data/xnprd/xnprddata/undo01.dbf' size 2000M autoextend off;

Tablespace created.

6:关掉数据库,取消掉pfile里的注释,并把undo_tablespace指向APPS_UNDOTS2

7:重启数据库,一切OK

来自 “ ITPUB博客 ” ,链接:http://blog.itpub.net/13561774/viewspace-573394/,如需转载,请注明出处,否则将追究法律责任。

请登录后发表评论 登录
全部评论

注册时间:2008-11-29

  • 博文量
    20
  • 访问量
    41062