How full restore is done using the SNAPSHOT api if snapshot is done incrementally?
Is NAS recommended for the backup repository or it can be any FS NFS mount?
A snapshot being incremental means that it will only copy the segments that have changed, for the segments that have not changed, a pointer to the old (already backed up) segments will suffice.
For a full backup then, it means copying the unique segments from a snapshot as well as resolving the pointers to any other segment files when copying the data from the repository back to the index location.
Is NAS recommended for the backup repository or it can be any FS NFS mount?
NAS is recommended only because it is an easy way to have a mount point that is accessible from all nodes in the cluster.