Storage Scale and Spectrum Archive: If a pre-migrated or migrated file is deleted, provide a limited window to recover it without having to use Storage Protect

Current documentation states that for a space-managed fileset if you use snapshots it triggers a recall of the data from tape to disk. The University of Michigan was always told this is a limitation in Storage Scale not Spectrum Archive and that there was work being done to address this.

Our use case for snaps on archive is exactly what you would think.  If someone deletes a pre-migrated or migrated file we want a limited window in which we could recover it without having to use Storage Protect.  (We don't have TSM on this system at all only replication).

We use snaps throughout our environment and our archive is meant to be our system of last resort and over time hold the most data, but an errant rm or ransomware/bad actor, then we could lose everything.

The expected behavior is that when you rm a pre-migrated file it could either move as is (still taking up disk space for the snap) or optionally stub to migrated status moving the inode to the snapshot leaving data on tape as valid.

For migrated data the inode would move and the data on tape would not show up as reclaimable etc. until the snap (all inodes referring to the entry on tape)  expire out the snaps.

This obviously covers the deleted case. We would hope for similar behavior for modifying write etc.

Short is accessing SA managed files in a snap or not in a snap would be the same.