I have a USB device that i connect to PC01.example.com that must reach serverA.example.com with an https get call. Todo that i installed the stable version of nginx on PC01 and i configured the server like this:

events {}

http {
    include       mime.types;
    default_type  application/octet-stream;

    server {
        listen 80;
        server_name myserver_80;
        return 301 https://$host$request_uri;  # Redirect HTTP a HTTPS

        location / {
            root   C:/nginx-1.26.1/html;
            index  index.html index.htm;
        }
    }

    server {
        listen 443 ssl;
        server_name myserver_443;

        ssl_certificate C:/nginx-1.26.1/ssl/certificate.crt;
        ssl_certificate_key C:/nginx-1.26.1/ssl/private.key;

        location /myapp{
            proxy_pass https://backend_upstream;

            proxy_set_header Host $host;
            proxy_set_header X-Real-IP $remote_addr;
            proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
            proxy_set_header X-Forwarded-Proto $scheme;

            proxy_pass_header Authorization;

            # Timeout settings
            proxy_connect_timeout 60s;
            proxy_send_timeout 60s;
            proxy_read_timeout 60s;
        }
    }

    upstream backend_upstream {
        server serverA.example.com;
    }
}

Now, the address serverA.example.com really is a load balancer with ip ‘203.0.113.2’ and that’s the problem:

when i call https://serverA.example.com/myapp translates the address to https://203.0.113.2/myapp and of course the call fails with a 502 Bad Gateway error because there is no web service on the load balancer at 203.0.113.2.

Is there a way to force nginx to use the fqdn for the upstream? Or, there is any other configuration i can change to make it work?

Some notes:

1. nginx certificates are self signed
2. the real ip of serverA.example.com is on another subnet and it’s not reachable from PC01 without using the load balancer.
3. if i call curl -k https://serverA.example.com/myapp/health from PC01 it works fine
4. if i call curl -k https://203.0.113.2/myapp/health from PC01 it doesn’t work
5. if i call curl -k https://localhost/myapp/health from PC01 it doesn’t work (of course)

I thought about adding a resolver to the configuration, but i am not sure what should i set it to.

i have tried various configurations, for example changing

proxy_set_header Host $host;

to

proxy_set_header Host serverA.example.com;

but nothing changed