How Must the Data Policy Avoid Overlapping Branch IPs?

Why the Answer Is Correct

Option A is the data-policy form that matches the overlapping branch source and invokes NAT so the remote site or server sees a unique translated address. In Cisco Catalyst SD-WAN, overlapping IPs cannot be resolved by routing alone because both branches advertise the same prefix; the data policy must intercept the matching traffic and apply a NAT pool action. The NAT pool gives each branch a distinct outside address, and return traffic is translated back before it reaches the original host. As Loi2525 noted, the intent is to translate the overlapping IP so it reaches the other side via an outside/global address.

Why the Other Options Are Wrong

Option B is incorrect because its interface address and policy source address do not align with the overlapping flow that must be matched; as jstaruch noted, there is a mismatch there. Option C shows an interface natpool/static-NAT snippet, which may define translation resources but does not by itself show the data-policy sequence that matches the overlapping traffic and calls the NAT action. Option D is a Direct Internet Access construct using nat use-vpn 0; it sends traffic locally to the internet and does not solve branch-to-branch or branch-to-server overlapping addressing.

Community Comment Notes

jstaruch correctly flagged the B mismatch and identified D as DIA rather than a NAT-pool solution. Loi2525 backed A because the overlapping address must be translated to an outside global address before the remote side can reach it. BECAUSE said "C looks like the only answer," and westh4m1234 quoted the natpool/static snippet from C, but those comments focus on the interface-level NAT object rather than the data-policy action the question asks about.