We have oracle server version 8.0.5 & it is OLTP server & running in archive log mode. Yesterday we faced a serious problem of REDO file corruption while long running trans. ( Disk had hardware fault). After replacing HDD , we are unable to recover database up to the failure point though we had applied all archive logs.
Note: We have only two redo groups having two members.
Pls. suggest to recover up to last running trans.
well you had two members in the same corrupted disk?
You'll have to specify everything if you want help. Where are the redo files ? If you multiplexed correctly, you have a good redo somewhere. The idea behind multiplexing is to write to 2 locations that are safely on different disks or even different controllers.
It is unclear where you stand.
Wel ! members of redolog are multiplexed on four different disks. Other members of redolog are healthy and good.
I try to recover from healthy redolog member by standard
procedure but can't successed. I don't know the other step by step procedure to fix this problem.
Suppose members of redolog are not multiplexed then is there any way to fix this problem ?
Pls. mention all req. steps to recover the same.
Recovery: from Logs
Hi, 20th April 2001 11:12 hrs chennai
Still i dont see from your reply confirming you have redo log copy.
Any how you can see the scenarios as follows.
If redo log or corrupted recovery to the time of failure may not be possible.However no data is lost under the following circumstances.
I)The redo logs files lost are not current.
II)The redo log had been archived.
III)The database is properly configured with mirrored redo log files.
Suppose members of redolog are not multiplexed ?as per your doubt the answer is you can as the reasons mentioned above.
If not you cannot recover up to the last point then You have to follow this steps for that.
You have to do CANCEL based recovery.
1)Shutdown the DB.
2)from the valid backup mount the instance
3)Restore all datafiles from the most recent backup
4)after checking the last redo log available from the archives
identofy the log seq no and issue the command
svrmgr>recover database until cancel
5)Open the DB with Reset logs option
6)Now the last data from the last commit has to be reentered.
a cpmplete recovery operation fails due to a bad or missing archived log.Recovery can be only completed to a time in the past prior to applying the archived logs
Missing redo Logs:
If the redo logs are not mirrored and you lost a redo log before it is archived,along with a datafile recovery cannot continue past the redo log.
Attitude:Attack every problem with enthusiasam ...as if your survival depends upon it
Click Here to Expand Forum to Full Width