First things first:
- 188.8.131.52 2-node RAC
- solaris 8 / sun cluster 3.0
- VxVM 3.2
- production system, rather urgent
we have this little situation where someone dropped a datafile using
'offline drop'. This was apparently not the best solution for its
problem and we don't have archivelogs available to recover the datafile
(only the archivelogs created during the backup are kept .. don't ask ..).
As I understand it right now, there's no way we can "undrop" this
datafile, our only option is to drop the tablespace and recreate it,
which is what we'll do eventually. To get out of this situation we would
like to continue for a while with this datafile dropped, while adding
another datafile to make up for the lost space.
Now, the question is, is this a good idea ? My main concern is
backup/recovery, but any other hints would be highly appreciated.
PS the original problem:
- added a datafile
- the second node didn't have its symlink to the new vxvm volume
- adding the datafile worked, everything fine on node 1
- on node 2, got ORA-01157
on the new datafile
- created symlink but the instance did not pick up the datafile
- offline drop the datafile to bring the db into service again (while a
'alter system check datafiles' would have fixed it)
COM D MN A
Email : Koen.Van_Langenhove@(protected)
- - - - - - -
Unix IS user-friendly .., it's just quite picky about its friends.