You absolutely can use the same LIFs for both the primary-secondary SnapMirror as well as the NDAS copy to cloud. that's the way I have it set up in my lab - just the one set of IC lifs. So long as they have network connectivity to the AWS resources (specifically S3 bucket) and are in the default IP space, you're good
NDAS backs up volumes, based on an ONTAP snapshot like SnapMirror. You can then restore full volumes, or search the catalog for individual files or LUNs to restore. So if you restored a full volume or a LUN containing VMs, then these would be restored, but with the snapshot-level consistency.
I can't discuss potential for VM-granular backup/restore here, but it's definitely something we can follow up on through your NetApp account team.
Also, I should have earlier responded that NDAS is generally available. You can see more information and register for a free 90-day trial at
Yes it is snapshot-based - it uses SnapMirror from primary to secondary. So if you've already got local snapshot policy in place on your volumes, then there is no additional overhead. NDAS takes the daily labelled snapshot off the secondary and sends it to cloud. Your existing local volume snapshot policy on the volume will be used on the primary.
If you start to protect volumes with NDAS which don't have a local snapshot policy (e.g. if you'd disabled the default ONTAP snapshot policy for some reason), then NDAS applies a default policy which is the same as ONTAP's default policy - Hourly snapshot keep 6, daily snapshot keep 2, weekly snapshot keep 2. So those will consume space on the primary. As you know, ONTAP snapshots are space efficient so that only the changed blocks take up real space.
The NDAS interface in the Data Protection screen tells you what protection is currently in place for each volume - represented by the green shields.