my production database hang. when the OS restart the oracle datafile of the redo
file broken. i try to alter database open resetlogs, but the database's datafiles are not consistent. Unfortunatly, this database do not backup completely recently.
Solution
1)_ALLOW_RESETLOGS_CORRUPTION = TRUE
Adding this parameter in pfile and startup with it. this prameter force database to open
without current online redolog and datafiles are inconsistent.
2)recover database
alter database openresetlogs;
after doing this operation, i can get the ORA- error in alert file.
==================
ORA-00600: internal error code, arguments: [2256], [0], [3221225472], [5], [3880116460], [],
[], []
==================
the fourth parameter [5] shows that what level we should rollback.
sqlplus "/ as sysdba"
alter session set events '10015 trace name adjust_scn level 24';
level 24 is computed by the fourth parameter from the ORA-00600 error.
(5+1)*4=24
Remark: level 1 will rollback 1 billion ( 1024 * 1024 * 1024 )SCN.
we can query this parameter from this view.
select KSPPDESC from X$KSPPI where ksppinm='_allow_resetlogs_corruption';
阅读(1029) | 评论(0) | 转发(0) |