[Info-vax] Delete a file from a tape?

[Christoph Gartmann]

In article <4AD54599.3550F594 at spam.comcast.net>, David J Dachtera <djesys.no at spam.comcast.net> writes:
>Michael Kraemer wrote:
>> 
>> If it's an ANSI label tape and one places an additional file mark
>> after the second last file, the last file will be deleted, sort of.
>
>BACKUP tapes are ANSI labelled - HDR1 and HDR2 only.
>
>So, the option you suggest will work, but ...
>
>Christoph:
>
>What you need to do is use SET MAGTAPE/SKIP commands as needed to
>position the tape at the HDR1 record for the last "file" on the tape,
>then issue:
>
>$ SET MAGTAPE/END_OF_FILE ddcu:
>
>....to overwrite that last HDR1 label record with a tape mark leaving two
>adjacent tape marks which indicate "End of volume". This renders the
>remaining data on the tape inaccessible.
>
>Use DUMP/BLOCK=COUNT=1 after SET MAGTAPE/SKIP=FILE=n to verify the tape
>position, then use "SET MAGTAPE/SKIP=FILE=-1 ddcu:" to back the tape up
>and leave it positioned just past the preceding tape mark.
>
>For example, if there are three good savesets before the final
>incomplete one:
>
>$ MOUNT/FOREIGN ddcu:
>$ SET MAG/SKIP=FILE=1 ddcu:	! SKIP past the first set of HDR labels
>$ SET MAG/SKIP=FILE=1 ddcu:	! SKIP past the first saveset
>$ SET MAG/SKIP=FILE=1 ddcu:	! SKIP past the first set of EOF labels
>$ SET MAG/SKIP=FILE=1 ddcu:	! SKIP past the 2nd set of HDR labels
>$ SET MAG/SKIP=FILE=1 ddcu:	! SKIP past the 2nd saveset
>$ SET MAG/SKIP=FILE=1 ddcu:	! SKIP past the 2nd set of EOF labels
>$ SET MAG/SKIP=FILE=1 ddcu:	! SKIP past the 3rd set of HDR labels
>$ SET MAG/SKIP=FILE=1 ddcu:	! SKIP past the 3rd saveset
>$ SET MAG/SKIP=FILE=1 ddcu:	! SKIP past the 3rd set of EOF labels
>
>....or...
>
>$ SET MAG/SKIP=FILE=9 ddcu:
>
>Use DUMP to verify that you're now sitting at the final set of HDR
>records:
>
>$ DUMP/BLOCK=COUNT=1 ddcu:
>
>(The tape is MOUNTed /FOREIGN. So, /BLOCK is needed rather than
>/RECORD.)
>
>The record displayed should begin with "HDR1" in the first four bytes
>and the saveset name should begin immediately thereafter in the record.
>If it does, you're o.k. Now, back the tape up over the record you just
>read:
>
>$ SET MAG/SKIP=FILE=-1
>
>...., then issue:
>
>$ SET MAG/END ddcu:
>
>The tape should now end with two adjacent tape marks. You can verify
>this with:
>
>$ SET MAG/SKIP=-3
                ^
                FILE=

>$ DUMP ddcu:
>
>....and you should see two end of file marks go by at the end of the data
>(reading will begin in the EOF records for the third good saveset)
>before the command completes successfully.
>
>$ DISM/UNL ddcu:
>
>I'm doing this from memory and cannot actually simulate the condition to
>verify that this will correct it. I'd need a good eight hours or more to
>set it up and verify it.
>
>Hope this helps. 

It did!

Thank you very much
   Christoph Gartmann

-- 
 Max-Planck-Institut fuer      Phone   : +49-761-5108-464   Fax: -80464
 Immunbiologie
 Postfach 1169                 Internet: gartmann at immunbio dot mpg dot de
 D-79011  Freiburg, Germany
               http://www.immunbio.mpg.de/home/menue.html