Re: Urgent RMAN Recovery question

"Richard S." <rshome_at_ntlworld.com> wrote in message news:ldnI9.634$iz3.70513_at_newsfep2-win.server.ntli.net...
> This is the scenario..
>
> Oracle 8.1.6 Enterprise
> Hot backups taken daily using RMAN script something like as follows...
>
> run
> { (allocate channels to disk}
> backup database including current controlfile;
>
> sql "alter system archive log current";
> sql "alter system switch logfile;"
>
> sql "alter database backup controlfile to ..."
> sql "alter database backup standby controlfile to ....";
> }
>
> We don't backup the archivelogs because we have a enormous amount of daily
> churn on the database and not enough space.
>
>
> We have tape restored the backup set files onto a different test server
for
> the purposes of trying the restore out. The backup was from over a week
ago,
> so the there are more recent backup set records in the RMAN catalog.
>
> I can succesfully RESTORE the database, but when I try to mount it in
every
> which way, I always end up getting ORA-01152 (file was not restored from a
> sufficiently old backup).
>

As you probably know, the 1152 error message usually occurs because an attempt has been made to perform an incomplete recovery, but not enough redo was applied, so one (or more) data files are actually sitting at a time in front of where the rest of the files have been recovered to. (And, incidentally, the error message usually indicates *which* file it is that is causing the problem. Knowing that might be of help).

I see no mention in your post of precisely what you've restored and how you did the restore. I also don't see any mention of what recovery commands you issue once the restore is complete.

You say that 'I successfully restore the database', but the error message indicates that it wasn't successful at all -that bits and pieces of whatever it is that you restored are from different times (ie, with different SCNs). So knowing how you do the restore is pretty important if anyone is going to be able to help you.

Likewise, without seeing what recovery commands you issue, it's impossible to provide any meaningful assistance.

The only thing I can say right now is that your script is no way to be backing up the controlfile, and it's almost certainly this fact alone that is causing you grief. You've already backed up the controlfile with the database (that's what the line 'including curent controlfile' does) so why you then have a 'backup controlfile to....' I've no idea. Using the second controlfile backup to perform restore/recovery is going to stuff you up big time, I suspect.

> We have a standby database which I've rebuilt several times using the
> backup, but have always created a standby controlfile from the primary at
> the time of recreation. This always works.
>
> However I'm trying to prove that the RMAN backup pieces, together with the
> RMAN catalog are sufficient in restoring and recovering without having to
> rely on creating a standby controlfile from the primary system.
>

The standby controlfile is a red herring. You really don't want to go down this route, because it means your recovered database would never be able to be opened properly. As the name implies, it would mean you'd just created a standby database, not a 'real' one.

> Can anyone please assist, fairly urgently.
>

Post further details, and I'm sure we can help.

Regards
HJR
> R.
>
>
>
Received on Sat Dec 07 2002 - 14:20:23 CST