Given that devices used in the backup are accessed using fully independent paths through the fabric and ports at the array. Which backup strategy has the least impact on perform the most improved availability to a production application running in an active/standby cluster?

A.
backup at the standby host from a split-mirror point-in-time copy.

B.
backup ata the primary host from a snapshot of a clone point-in-time copy

C.
backup a the standby host from a delta-snapshot of the production disks

D.
backup at the primary host from a multi-point snapshot of the production disks