You are using 'Distributed Tracing' to debug latency. The trace starts at the Gateway but 'dies' before reaching the backend; the backend logs show a new Trace ID. What is the standard Gateway configuration required to fix this?

1. A The Gateway must enable 'Packet Inspection' to find the Trace ID buried deep in the TCP segment payload.
2. B The Gateway must be configured to 'Propagate Context Headers' (like W3C `traceparent` or Zipkin `X-B3-TraceId`) from the incoming client request to the outgoing upstream request.
3. C The Gateway must share its internal CPU clock with the backend services via NTP to ensure ID synchronization.
4. D Distributed tracing only works if the Gateway and all backends share the same physical server.