I can't recall this ever happening before, on a test database (9ir2),
users wanted a fixed point in time to go back to, so I do a clean
shutdown, backup all datafiles (yes, redolog and control files too) to
another drive, restart database, which is not in archivelog mode. About
an hour later, user comes over and they want to go back to the point in
time it was last backed up. No problem, shut the db down, delete the
datafiles, control files and redolog files and copy the ones from the
backup to their original destinations. Now I have the database *exactly*
as it was, as I have done many times before.
When I start up the database it tells me that the system tablespace is
in need of recovery. Why?, should the database not just startup just as
if I did a shutdown/startup? I ended up mimicking a incomplete recovery
using recover database until cancel and open with resetlogs, but this
should not have been neccesary, should it?
Received on Fri May 20 2005 - 16:57:57 CDT
