On 6/27/06, Joe Armstrong-Champ <joseph.armstrong-champ@xxxxxxxxx> wrote:
Jared,
What about the use of the catalog in disaster recovery? We have ours on a seperate server from the prod databases so if the prod server fails we can still recover. Although as I'm writing this I'm not sure if we need the catalog since we would have the backup on tape (including the controlfile autobackup) and could still recover with that.
Joe
That might be a valid reason for having a recovery catalog, it you need it for disaster recovery.
If you are in the habit of having at least 3 controlfiles per database, all on physically separate volumes, the risk of needing to restore without a controlfile is fairly low.
What could cause that scenario?
A disaster is one way that could happen, as you mention. But if you have a disaster recovery site, then your database(s) would be replicated in some manner. In the event of a real disaster, recovering your database at the original site might not be your 1st priority. Running for your life might top the to-do list that day.
All the volumes with controfiles on them could be corrupted. Unlikely, but possible.
Someone deletes all the controlfiles. That's another possibility. More likely than the HW failing IMO.
AAMOF, I know of a recent case of that happening at a rather high profile site. This company could not ship *any* product for most of a day, because a DBA deleted all the database files for a rather important app.
Oops, wrong server. :(
-- Jared Still Certifiable Oracle DBA and Part Time Perl Evangelist
