At a disaster planning meeting, an architect is presented with the following facts by a client:
A Unix host is replicating storage from a Production PowerMax to a VMAX 250FX.
RDF pairs have been split with Production and DR hosts accessing their respective storage.
Invalid tracks have started to build against PowerMax and VMAX 250FX.
The client now wants to bring production back to PowerMax, discarding the VMAX 250FX updates while keeping the updates on the PowerMax. What is the correct action to take?