rac断电后打开提示datafile 1 有坏块，recovery stuck at block n of file 1

其实只在虚机上搭过rac，没在真实环境中碰到过，机会来了。。

内网两台RAC，断电后无法启动。

过程：（前后持续几个月。开始看了两天没成功，后面就没管）

1、alert报错

Tue Jul 31 14:16:25 2018

Hex dump of (file 1, block 1824271) in trace file /oracle/ora11g/diag/rdbms/rac/rac4/trace/rac4_p001_8985.trc

Tue Jul 31 14:16:25 2018

Hex dump of (file 1, block 1824272) in trace file /oracle/ora11g/diag/rdbms/rac/rac4/trace/rac4_p006_8995.trc

Reading datafile '+DATA/rac/datafile/system.256.856700057' for corruption at rdba: 0x005bd610 (file 1, block 1824272)

Reading datafile '+DATA/rac/datafile/system.256.856700057' for corruption at rdba: 0x005bd60f (file 1, block 1824271)

Reread (file 1, block 1824271) found same corrupt data (logically corrupt)

RECOVERY OF THREAD 4 STUCK AT BLOCK 1824271 OF FILE 1

Reread (file 1, block 1824272) found same corrupt data (logically corrupt)

RECOVERY OF THREAD 4 STUCK AT BLOCK 1824272 OF FILE 1

执行了一下   recover database using backup controlfile until cancel;

提示ora-1547,ora-1194

2、设置隐藏参数 _allow_resetlogs_corruption=true, _corrupted_rollback_segments="xx"    undo_management=manual

alter database open; 提示错误：

ora-600  [kclchkblk_4]

ora-704: bootstrap process failure

3、推进scn     oradebug poke  或 alter sesession set events '10015 trace name adjust_scn level x' (11.2.0.4 之前可能才有效）

4、alter database open;  成功，并未需要resetlogs，可能和此处理方法有关。

5、alert日志，大量ora-600 [ktbsdp1] ，ora-01555, ora-600 [4097]  似乎都和undo损坏有关

6、删除当前undo表空间 重建

7、取消隐藏参数，重启

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/24680677/viewspace-2564640/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/24680677/viewspace-2564640/