Best possible solution to reduce the restoration window for Oracle DB

Yes you can restore an oracle database taken with a hardware snapshot and restore to dell but not with the performance benefits you are looking for. The process would be:

1. Snap application
2. Backup copy the data out of the snapshot to traditional media (e.g HSX disk)
3. (optional) copy the data to a storage target closer to the dell
4. Traditional restore back to the dell (this is a traditional disk library → server restore)

This helps accelerate protection but not recovery since you are performing the same type of restore in the end. Accelerating recovery with snapshots means reverting the snapshot (i.e overwriting production with another snapshot on the array) which is not possible out-of-place in your scenario.

Restoring 100GB in 25 hours is horrible performance at 4GB/hr. It sounds like something more fundamental is wrong with the infrastructure - either there is a network limitation/bottleneck or source disk read limitation on the HSX cluster, which seems unlikely, OR that dell disk is really slow.

I think in your case you should do large file copy tests to figure out if the bottleneck is the source, network or destination target (or a combo of both).

Since it sounds like you are already replicating the file system, have you considered replicating the database as well?