I think to give you advice we need more details. Is that extra disk a backup of your SATA root/boot disk, or of one of the SAS data disks in the RAID? What do you mean by "recover"? Has your current disk failed and you want to substitute the backup for it? Do you want to mount the backup disk (while the system continues to operate normally) and pull a few files from it? Have you modified the primary disks since making the backup?
In general, using dd to copy a whole disk (or file system partition) is no longer considered a good way to do backups. The reason is exactly what you hinted at: In modern file systems (such as ZFS), the data on disk includes information about what this disk contains (which pool, which file system, ...), and if the OS sees two copies of the same information, it is easy to get confused. For traditional file systems (such as UFS and ext2), the dd technique worked decently, because the information about which partition is used how and where is not in the partition itself, but in the fstab file in /etc.
Today, to take backups of a ZFS file system, the most accepted technique is to use send/receive to a second disk, and organize the data by snapshots.
