Help? Unable to access the UI with a reverse proxy.

[Nedevski]

Hi, I'm unable to access the UI with my reverse proxy.

I have a pretty basic setup - a container with two networks - the default one and one for Nginx Proxy Manager.
There I create a basic binding for http://ddns-updater:8000/ and give it a URL. This doesn't work.

However when I go to http://server_up:18000/ everything loads perfectly fine.
All my other services are set up this way and it works fine, but not in this case.

Is there any additional setup that I need to do?
Here is my docker-compose.yml file:

version: "2.1"

services:
  ddns-updater:
    image: qmcgaw/ddns-updater
    container_name: ddns-updater
    volumes:
      - ./volume:/updater/data
    networks:
      - default
      - nginx
    ports:
      - 18000:8000
    restart: unless-stopped

networks:
  default:
    name: ddns_updater_default
  nginx:
    name: nginx_default
    external: true

[LeVraiRoiDHyrule]

Did you try to ping ddns-updater container from nginx container ? Are you sure it is reachable ? I proxy the web UI using Caddy with no particular setting. You don't have any https ?

[qdm12]

This is likely an nginx configuration issue, are you sure you are proxying to port 8000 of the container and not 18000?
Try exec-ing in the nginx container and see if you can resolve ddns-updater correctly etc. i.e.

docker exec nginx wget -qO- http://ddns-updater:8000

[Nedevski]

First - clarification, it's not nginx, it's nginx-proxy-manager, which I guess is the most popular UI manager for nginx.

And yes, I'm sure I'm exposing the container port. Both containers are in the same network, the domain mapping is for http://ddns-updater:8000, Chrome returns DNS_PROBE_FINISHED_NXDOMAIN.

However when I try with the http://server_up:18000/ address - it works.
I have 20-ish other services and the mapping for them works as intended, not sure what the issue is here. I will try to ping the container from docker console later when I have access to the server.

[qdm12]

DNS_PROBE_FINISHED_NXDOMAIN means the domain name ddns-updater doesn't exist. Are you sure they're part of the same docker network, because it looks like the docker local dns part of the docker network of nginx doesn't see ddns-updater

[anandslab]

I would like to second this issue. I am facing the exact same issue but with Traefik reverse proxy. I can access via http://host-ip:8000 but I cannot access via Traefik. And yes, ddns-updater is part of Traefik network.

[qdm12]

The ddns container is healthy right? By default traefik doesn't serve unhealthy containers, maybe that's the issue. Otherwise, try shelling into the traefik container and checking you can resolve the ddns container name from there, to ensure it's part of the same docker network.

The page is a "dumb" static page (only html and css) so I doubt it's an issue with the webpage or http server really.