With using both servers actively I mean both of them can serve half of the underlying storage through iscsi. When one fails I have to be able to move the iscsi connections of the failing node to the remaining node.