I'm sorry, but you cannot recover to any point in time after the gap. I would export what you can from the good tablespaces. Get a copy of BMC LogMaster (or whatever it is called now) to dump out the available redo since the gap. Recover the database to just before the gap. Attempt to run the SQL extracted from the log files against the recovered database. Drop and import the exported tables. Then do a painstaking check to find the extent of the damage.
Because of various backup problems that would take days to explain, we have need to recover a database that is missing one day's worth of archive log files. This 7.3.4.5.0 database on Solaris 2.6 had all of its tablespaces put into hot backup mode at once 7 days ago and for various reasons, were never taken out of backup mode. Last night, we had a crash during the nightly hot backup and lost two filesystems. We have restored the two filesystems, /s009 and /s008, one from last night's backup and one from the night before's backup. That brought us to a state of having a mounted database with all of its datafiles in backup mode. We did alter database datafile 'name' end backup on all of the datafiles, except the ones in /s008 and /s009 (see below). For the ones in the restored filesystems, we received an Oracle error saying that the files are old versions:
SVRMGR> select name from v$datafile a, v$backup b
2> where a.file# = b.file#
3> and b.status='ACTIVE';
NAME
--------------------------------------------------------------------------------
/s008/oradata/PROD/fin.dbf
/s008/oradata/PROD/mfgx.dbf
/s008/oradata/PROD/mfgx02.dbf
/s008/oradata/PROD/mfgx03.dbf
/s008/oradata/PROD/fin02.dbf
/s008/oradata/PROD/mfgx04.dbf
/s008/oradata/PROD/mfgx05.dbf
/s009/oradata/PROD/rbs_mnstr_f04.dbf
8 rows selected.
SVRMGR> alter database datafile '/s009/oradata/PROD/rbs_mnstr_f04.dbf' end backup;
alter database datafile '/s009/oradata/PROD/rbs_mnstr_f04.dbf' end backup
*
ORA-01235: END BACKUP failed for 1 file(s) and succeeded for 0
ORA-01122: database file 56 failed verification check
ORA-01110: data file 56: '/s009/oradata/PROD/rbs_mnstr_f04.dbf'
ORA-01208: data file is an old version - not accessing current version
SVRMGR>
When we try to do recover datafile on this datafile
(rbs_mnstr_f04.dbf), Oracle says it needs an archive
log file from 7 days ago, which was the day that all
of the datafiles were put into backup mode and never
taken out of it. The problem with this is that 5 days
ago, due to human error, we lost (accidentally
deleted) a day's worth of archive log files. Since
Oracle is asking for archive log files from before the
"gap", we are trying to find out how to convince Oracle that it doesn't need those archive log files
since we have all of the datafiles from a point in
time after the gap and also all of the archive log
files after the gap. I have tried creating a
controlfile and also doing recover database using
backup controlfile and various other forms of recover
database/recover datafile and Oracle always asks for
arch logs from before the gap - the day that they were
put into hot backup mode. The only thing I've had
success with is simulating this scenario on a test
database in which I did recover database using backup
controlfile until cancel, which I canceled immediately
when asked for archlogs from before the gap. I then
was able to shut down the database, set
_allow_resetlogs_corruption=true in init.ora, and then
startup and open with resetlogs. I am assuming,
however, that I now have logical corruption, since
none of my archlogs from after the gap were applied to
the datafiles I just opened with that command.