Which practice helps avoid NAT double translation in an SD-WAN deployment?

• A. Use a single egress point per path and align NAT rules to avoid double translation
• B. Enable NAT on all hops to ensure coverage
• C. Use dynamic IP addresses for all devices
• D. Disable NAT entirely

Answer: (A)

Avoiding NAT double translation means ensuring traffic is translated only once, by a single egress point for each path, and keeping that translation consistent across the path. When NAT happens at multiple points along the path, the source IP can be rewritten more than once, which breaks connectivity and complicates return traffic. By using one egress point to perform NAT and aligning the NAT rules across all paths, you guarantee that outbound translations are uniform and that responses map back correctly, no matter which path is chosen.

If NAT were applied at every hop, you’d end up translating the source address multiple times, leading to double translation and connection issues. Simply relying on dynamic IPs or disabling NAT doesn’t address the root problem and can cause routing or reachability problems.