May I ask does this appear at your origin host / server logs or?
Yes, this shows up in the origin/upstream logs.
Furthermore, may I ask did you followed the instructions from below article or?:
No, I use Argo tunnels for proxying to the upstream.
This is most certainly a bug in how CF-Connecting-IP is being set - would be nice to get a confirmation (or not) from a Cloudflare staff member. I’m not sure if it only happens within Argo tunnels or not. Also, I’ve since stopped relying on the header so I don’t know if this is still happening currently.
backend image: node:14.15.1-alpine; app running on fastify v3
proxied through: nginx:1.17.9-alpine
Nothing had changed in the deployment environment - no new deployments or restarts. The header’s value just abruptly changed. It might also be worth noting that the first error we saw had the header value as an IPv6 address, also duplicated, and then subsequent errors had IPv4 addresses, duplicated.