Re: Applying archivelogs to a cold backup - shortcut?

Ok, this seems to have worked. I'm tail -f'ing the alert log, and it's working. Slowly, but working.

I guess this will be my reminder to automate the db backup very soon.

Thanks for you help Howard, you saved my ass.

    Jason

"Night Shade Books" <jasonw_at_nightshadebooks.com> wrote in message news:6E819C57D5B93060.9E19F01542E81E60.F3B05E34554F783D_at_lp.airnews.net...
> Hrmm. I'm not sure I understand.
>
> When I took a cold backup, I backed up the datafiles, the controlfiles,
and
> the redo logs.
>
> When I restored, I copied them all back to the appropriate directories.
> Then I went into svrmgrl and started it up. Worked like a charm. Except
it
> was out of date. So I wanted to apply the archive logs to bring it up to
> date.
>
> If I do a 'recover database' it tells me that recovery is not needed. So
> what I'm doing now is a
> 'recover automatic database using backup controlfile until cancel'. I'm
not
> sure if it's working or not, but it seems to be.
>
> Jason
>
> "Howard J. Rogers" <dba_at_hjrdba.com> wrote in message
> news:3bf8c674$0$13483$afc38c87_at_news.optusnet.com.au...
> > That's probably because you restored the control files (and online redo
> > logs?). You shouldn't have done. If performing an incomplete recovery
> > (which is what an 'until cancel' is), you *only* restore the datafiles
> (all
> > of 'em).
> >
> > If you restore the whole lot (especially from a cold backup) the thing
is
> > entirely consistent and requires no recovery.
> >
> > If you have genuinely lost all controlfiles and all redo logs, it's not
> just
> > an automated backup strategy that's up the spout, it's the entire
database
> > configuration. Whatever happened to multiplexing of control files and
> redo
> > logs?
> >
> > Regards
> > HJR
> > --
> > Resources for Oracle: http://www.hjrdba.com
> > ===============================
> >
> >
> > "Night Shade Books" <jasonw_at_nightshadebooks.com> wrote in message
> >

news:6D701DD75D0F8997.F91990E9A534E0B2.110CB707C27625D2_at_lp.airnews.net...
> > > Basically, the backup isn't automated for this db yet. Still working
on
> > > that.
> > >
> > > But I run into a problem doing 'recover database' or 'recover
automatic
> > > database'. It gives me an ORA-00264: no recovery required.
> > >
> > > Help?
> > >
> > > Jason
> > >
> > > "Howard J. Rogers" <dba_at_hjrdba.com> wrote in message
> > > news:3bf8c2f8$0$13483$afc38c87_at_news.optusnet.com.au...
> > > > Oops, that will teach me to slow down and read more carefully.
> > > >
> > > > You're doing an incomplete recovery (...until cancel) and you're
using
> a
> > > > backup controlfile. Both will require you to issue an 'alter
database
> > > open
> > > > resetlogs' when all the available archives have been applied. That
in
> > > turn
> > > > requires you (if you've got any sense) to then perform an immediate
> > > > shutdown, followed by a new closed database backup (resetlogs
renders
> > all
> > > > prior backups and archives useless).
> > > >
> > > > If you follow the advice in my earlier message, and issue the 'auto'
> > > command
> > > > when first prompted for an archive to apply (or the 'recover
automatic
> > > > database until cancel') then the first run through will apply, let
us
> > say,
> > > > 800 logs -and then bomb out, because the 801st log (presumably the
> > current
> > > > redo log) is missing (I presume that's why you're doing an 'until
> > cancel'
> > > > recovery anyway.... let's hope so). That's OK: when the recovery
> > process
> > > > fails to locate log 801, it will crash out horribly: you just start
it
> > up
> > > > again, issue the 'recover database until cancel' command once more,
> and
> > > this
> > > > time the first log it prompts for will be the missing one -so *now*
> you
> > > type
> > > > 'cancel', and that terminates recovery. Next comes your alter
> database
> > > open
> > > > resetlogs, and then the shutdown and new backup.
> > > >
> > > > Excuse me for asking, though: what kind of a database is it that
> forces
> > > you
> > > > to restore a backup from over 800 logs ago? Either you are backing
up
> > > every
> > > > night (in which case 800 log switches implies your redo logs are
> > > hopelessly
> > > > small), or you are backing up once in a blue moon -in which case,
one
> > > > wonders whether a slightly higher frequency of backups might be in
> > order!
> > > >
> > > > Regards
> > > > HJR
> > >
> > >
> >
> >
>
>
Received on Mon Nov 19 2001 - 03:09:13 CST