日期:2014-05-16 浏览次数:20530 次
今天以前某客户联系我,说有个库无法启动,花了几分钟远程看了一下,alert log信息如下:
Sun Mar 16 20:27:49 2014
Media Recovery Start
parallel recovery started with 7 processes
Sun Mar 16 20:27:49 2014
Recovery of Online Redo Log: Thread 1 Group 3 Seq 7363 Reading mem 0
Mem# 0 errs 0: /oracle/data/orclzjs/redo03.log
Sun Mar 16 20:28:01 2014
Errors in file /opt/app/oracle/admin/orclzjs/udump/orclzjs_ora_21469.trc:
ORA-00333: 重做日志读取块 67221 计数 6144 出错
ORA-00312: 联机日志 3 线程 1: '/oracle/data/orclzjs/redo03.log'
ORA-27072: 文件 I/O 错误
Linux-x86_64 Error: 2: No such file or directory
Additional information: 4
Additional information: 67221
Additional information: 669184
Sun Mar 16 20:28:08 2014
Errors in file /opt/app/oracle/admin/orclzjs/udump/orclzjs_ora_21469.trc:
ORA-00333: 重做日志读取块 65173 计数 8192 出错
ORA-00312: 联机日志 3 线程 1: '/oracle/data/orclzjs/redo03.log'
ORA-27091: 无法将 I/O 排队
ORA-27072: 文件 I/O 错误
Linux-x86_64 Error: 2: No such file or directory
Additional information: 4
Additional information: 67221
Additional information: 669184
Sun Mar 16 20:28:19 2014
Media Recovery failed with error 333
ORA-283 signalled during: ALTER DATABASE RECOVER database ...
Sun Mar 16 20:30:16 2014
ALTER DATABASE RECOVER database until cancel
Sun Mar 16 20:30:16 2014
这是很典型的Oracle currnet redo logfile损坏的案例。 询问发现该库是强制关的,重启主机后发现数据库无法正常启动了。
对于非归档的数据库,在停库之前,我们建议先进行alter system checkpoint,多切几次redo,然后kill 掉相关进程后再去shutdown。 哪怕是shutdown abort,这样风险