由于服务器断电导致数据库启动报错
|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Mon Sep 07 12:47:52 2026 ALTER DATABASE OPEN Beginning crash recovery of 1 threads ``parallel recovery started with 32 processes Started redo scan ORA-00354: 损坏重做日志块标头 ORA-00353: 日志损坏接近块 41735 更改 2385313823 时间 09``/07/2026 10:35:04 ORA-00312: 联机日志 5 线程 1: ``'D:\APP\LIMSADMIN\ORADATA\ORCL\REDO05.LOG' Mon Sep 07 12:47:58 2026 Trace dumping is performing ``id``=[cdmp_20260907124758] Errors ``in file d:\app\limsadmin\diag\rdbms\primary\orcl\trace\orcl_ora_3040.trc: ORA-00354: 损坏重做日志块标头 ORA-00353: 日志损坏接近块 41735 更改 2385313823 时间 09``/07/2026 10:35:04 ORA-00312: 联机日志 5 线程 1: ``'D:\APP\LIMSADMIN\ORADATA\ORCL\REDO05.LOG' ORA-354 signalled during: ALTER DATABASE OPEN... |
从这里看,数据库在启动的时候实例恢复需要读取REDO05.LOG日志,但是发现该日志中有block损坏,而且损坏日志的时间点为09/07/2026 10:35:04,scn为2385313823（ORA-00353: 日志损坏接近块 41735 更改 2385313823 时间 09/07/2026 10:35:04）,导致实例恢复无法完成,从而数据库无法正常启动.通过Oracle Recovery Check脚本检查数据库当前状态,数据库的checkpoint time为:09/07/2026 10:37:37,checkpoint scn为:2385314357
这个就比较明显,数据库文件正在需要的日志是sequence 19001 REDO01.LOG(而不是报错的sequence 19000 REDO05.LOG),基于这样的情况,出现这种问题,是由于control中记录的Thread Checkpoint RBA不正确导致.解决这个问题相对比较简单,重建ctl即可
由于redo05.log有损坏,该库为归档模式,需要先clear redo05.log（不然会导致redo无法归档,数据库hang住）,然后对全库发起备份