>    I would like to hear also...
>    Would it not be better to set the tape to "DESTROYED" ?
>    For some reason that term, in my mind, tends to indicate that
>    some action would be triggered when ADSM would notice a tape
>    volume in that state...
>    I thought I remember reading where if a copy is lost (being a most
>    recent) it would automatically pull in a new copy during next
>    incremental but can't find it now in my manuals...
>    ALSO BY DELETING YOU NOW HAVE NO MANUAL METHOD OF SEEING WHAT WAS ON
>    THAT TAPE... IF YOU MARK DESTROYED YOU CAN STILL DO A QUERY CONTENTS
>    later
>         Dwight


The primary reason for marking the volume destroyed would be so files
on the volume could be restored from copy storage pools.  Marking a
volume as destroyed would not cause files on the volume to be backed
up again during the client's next incremental backup.

Deleting the primary volume with discarddata=yes, would removed *all*
database references to the file versions that were on the tape.  This
means that the database would have no record of these bitfiles in any
storage pool.  This would cause these files to be considered for backup
during the client's next incremental backup.  However, it would not
be possible to restore files from copy storage pools, since the database
references to these bitfiles would no longer exist.

Dave Cannon
ADSM Server Development