RE: database recovery help
- From: "Peter McLarty" <p.mclarty@xxxxxxxxxx>
- To: <godwin.ror@xxxxxxxxx>, <oracle-l@xxxxxxxxxxxxx>
- Date: Tue, 6 Mar 2007 16:03:03 +1000
Hi Godwin
You mentioned that your backup is a hot backup and then you mention
doing a copy (cp) to restore those. How did you do the backup.
You will need the control file from your backup as it will contain an
SCN number that matches the SCN number for the datafiles you backed up.
Do you know when the corruption occured and how do you know its corrupt?
If your hot backup is ok then and you can pinpoint the time the
corruption occurred then it may be possible to do a recover until just
before the corruption occurred so not losing as much as the five days
work.
The recover database until cancelled will look in the archive log
directory as defined in your init.ora/spfile
To give you some better advice we need to know the above.
Cheers
--
Peter McLarty
Database Administrator
Student System Upgrade Project
Central Queensland University
________________________________
From: Godwin vincent [mailto:godwin.ror@xxxxxxxxx]
Sent: Tuesday, 6 March 2007 03:40 PM
To: oracle-l@xxxxxxxxxxxxx
Subject: database recovery help
Hi all,
Something went wrong and now i have corruption in my database
and it wont start at all. All I have is a backup which is 5 days old. As
part of the backup, i have a hot backup of all the datafiles, control
file and archive redo log files, which are all 5 days old. Now, i want
to recover my database using this backup (even though i loose a week
worth of data). I have never done a recovery ( either user-managed or
using RMAN) and would like to request some help in this regard. The
database is able to mount but is is not able to open, so im assuming
that there is no problem with the control file. Im running Oracle 8 on
Solaris. So, from my understanding, i have to perform a restore and
recovery and to do that will the following work? and is this the
standard way of restore and recovery
1. Restore the datafiles backup.
cp /u08/backup/datafilebackup1.dbf /u02/oradata/DB1/datafile1.dbf
cp /u08/backup/datafilebackup2.dbf /u04/oradata/DB1/datafile2.dbf
............... restore all the datafiles backups to the original
datafile location..........
2. startup mount
SQL> startup mount;
3. Recover
SQL> recover database until cancel;
When prompted give the path and file name of the archived
redo log to apply until it gives
"Media Recovery Complete", if it still asks for archived
logs when there are no archive
logs left to give, then give the location of the latest
active online redo log file.
4. open the database
SQL> aler database open resetlogs;
Is the above process standard way of recovering a database with the
backup?
1.Im still not clear on this control file, In this case, do i have to
restore the 5 days old control file to recover the database? or can i
use the current control file ( because there is no problem with the
current control file as the database is mounting)?
2. Where does the RECOVER DATABASE UNTIL CANCEL command look for
archived redo log files by default? can i restore the backup arch logs
to the archive log destination?
Any information would be of a great help.
Thanks,
Godwin.
Other related posts:
