I assume we are talking 7-mode here. Are you planning on use Oncommand to manage datasets for snapvault? If so, then you need qtrees. If you aren't planning on using OCUM, then you can do volume based via the command line. But it will be volume -> qtree vol/vol1 -> /vol/dst/qtree
I am planning on using OCUM. I have created the snapvault jobs from here as well and plan to create flexclone of the volume as a way to restore to an earlier snapvault initiated snapshot.
I thought you didnt need qtrees for flexvolumes whch are just used as data luns? there is an option to create a none qtree snapvault. I have had to create a qtree on the destination end which isnt an issue.