Re: Recovery from Hot Backup

<laslettm_at_my-deja.com> wrote in message news:95isd5$c4r$1_at_nnrp1.deja.com...
> I require clarification on restoring a database from a hot backup, using
> a backup controlfile.
>
> By way of background:
>
> I am running Oracle 8.0.5 (part of the one-hour install of Applications
> 11.0.3)
> Solaris 2.6
>
> After some problems applying patches to a production & test databases, I
> have installed 1103 on a seperate server, and am trying to copy the
> database from the original server to the "new" server.
>
> I took the nightly backup tape, shutdown the database on the new server,
> and restored all the datafiles.
>
> I have edited the backupcontrolfile script and run it so that the
> instance has started, and the db requires recovery.
>
> Problem:
>
> I have tried the following:
>
> recover database until cancel using backup controlfile;
>
> and
>
> recover database until time <time of an appropriate log switch from the
> alert.log> using backup controlfile;
>
> and keep getting message -
>
> ORA-01547: warning: RECOVER succeeded but OPEN RESETLOGS would get error
> below
> ORA-01194: file 1 needs more recovery to be consistent
> ORA-01110: data file 1: '/u02/oradata/LOOKPRD/system01.dbf'
>
> The archived redo logs I am pointing to have been archived well after
> the backup completed (up to 2 days in some cases!).
>
> The data in the db is not particularly important, but I need to get the
> db open.
>
> Can anyone point out what I am doing wrong? Am I missing something to do
> with the ONLINE redo logs?

No, you are misusing the recover database command. You say you have reconstructed your controlfiles from a trace script? Well, then they're not backup controlfiles. That command ONLY applies when you restore binary versions of the control file from a backup, not when you've rebuilt the thing using a trace script.

Not quite sure, either, why you are recovering until cancel. Presumably, you have a gap in your archive sequence? Otherwise, a complete recovery should be possible (and is thoroughly recommended).

The command should be: recover database. Period.

Restore your datafiles again. Then re-run the create controlfile script (because that will use the timestamps in the headers of the datafiles to decide what to stick in the header of the new controlfile). Then do a complete recovery.

Regards
HJR
>
> Thanks,
> Martin
>
>
> Sent via Deja.com
> http://www.deja.com/
Received on Sun Feb 04 2001 - 00:47:03 CST