Need help routing other containers through the VPN (following the documentation steps)

[amrdeus2]

Hey all,
Apologies in advance if this is a stupid question. I got Transmission working fine, but when following the documentation to use "network_mode: "service:transmission-openvpn" I'm getting the following error from Portainer:

"failed to deploy a stack: Network arr_default Creating Network arr_default Created Container transmission-openvpn Creating Container transmission-openvpn Created Container prowlarr Creating Error response from daemon: conflicting options: port publishing and the container type network mode"

Any help would be greatly appreciated. ChatGPT says "The error message is indicating that you're trying to publish ports (ports property) for containers that are also using the network mode of another service (network_mode: "service:transmission-openvpn")." Then it directs me to do it differently to the documentation, which eventually gets it all installed but the containers don't open their webUIs.

Any help would be amazing!
Here is my compose (pre any AI help):

`version: '3.3'
services:
    transmission-openvpn:
     container_name: transmission-openvpn
        cap_add:
            - NET_ADMIN
        volumes:
        - /volume1/docker/ipvanish/:/etc/openvpn/custom/
        - /media/USB/downloads:/downloads #optional
        - /media/USB/data/transmission:/config #optional?
        environment:
        - OPENVPN_PROVIDER=custom
        - OPENVPN_CONFIG=bulletVPN-AU-Sydney_1-syd01-tcp
        - OPENVPN_USERNAME=XXXXX
        - OPENVPN_PASSWORD=XXXXX
        - LOCAL_NETWORK=192.168.0.0/16
        - OPENVPN_OPTS=--inactive 3600 --ping 10 --ping-exit 60
        logging:
            driver: json-file
            options:
                max-size: 10m
        ports:
            - '9091:9091'
            - 8888:8888/tcp # HTTP proxy
            - '8989:8989' # port for sonarr
            - '7878:7878' # port for radarr
            - '9696:9696' # port for prowlarr
            - '6767:6767' # port for bazarr
        restart: always
        image: haugene/transmission-openvpn



version: "2.1"
services:
  radarr:
    image: lscr.io/linuxserver/radarr:latest
    container_name: radarr
    environment:
      - puid=998
      - pgid=100
      - TZ=Australia/Sydney
    volumes:
      - /media/USB/data/radarr:/config
      - /media/USB/data:/data #optional

    ports:
      - 7878:7878
    restart: unless-stopped
    network_mode: "service:transmission-openvpn" # Add to the transmission-openvpn Container Network
    depends_on:
      - transmission-openvpn # Set dependency on transmission-openvpn Container
    healthcheck: # Here you will check if transmission is reachable from the Jackett container via localhost
    test: curl -f http://localhost:9091 || exit 1
            # Use this test if you protect your transmission with a username and password
            # comment the test above and un-comment the line below.
            #test: curl -f http://${TRANSMISSION_RPC_USERNAME}:${TRANSMISSION_RPC_PASSWORD}@localhost:9091 || exit 1
   interval: 5m00s
   timeout: 10s
    retries: 2



version: "2.1"
services:
  sonarr:
    image: lscr.io/linuxserver/sonarr:latest
    container_name: sonarr
    environment:
      - puid=998
      - pgid=100
      - TZ=Australia/Sydney
    volumes:
    - /media/USB/data/radarr:/config
    - /media/USB/data:/data #optional
    ports:
      - 8989:8989
    restart: unless-stopped
    network_mode: "service:transmission-openvpn" # Add to the transmission-openvpn Container Network
    depends_on:
      - transmission-openvpn # Set dependency on transmission-openvpn Container
    healthcheck: # Here you will check if transmission is reachable from the Jackett container via localhost
    test: curl -f http://localhost:9091 || exit 1
            # Use this test if you protect your transmission with a username and password
            # comment the test above and un-comment the line below.
            #test: curl -f http://${TRANSMISSION_RPC_USERNAME}:${TRANSMISSION_RPC_PASSWORD}@localhost:9091 || exit 1
   interval: 5m00s
   timeout: 10s
    retries: 2



version: "2.1"
services:
  prowlarr:
    image: lscr.io/linuxserver/prowlarr:latest
    container_name: prowlarr
    environment:
      - puid=998
      - pgid=100
      - TZ=Australia/Sydney
    volumes:
      - /media/USB/data/prowlarr:/config
    ports:
      - 9696:9696
    restart: unless-stopped
    network_mode: "service:transmission-openvpn" # Add to the transmission-openvpn Container Network
    depends_on:
      - transmission-openvpn # Set dependency on transmission-openvpn Container
    healthcheck: # Here you will check if transmission is reachable from the Jackett container via localhost
    test: curl -f http://localhost:9091 || exit 1
            # Use this test if you protect your transmission with a username and password
            # comment the test above and un-comment the line below.
            #test: curl -f http://${TRANSMISSION_RPC_USERNAME}:${TRANSMISSION_RPC_PASSWORD}@localhost:9091 || exit 1
   interval: 5m00s
   timeout: 10s
    retries: 2
`

[pkishino]

If I recall correctly, you need to manually create a network (always recommended over the default anyway..)

[pkishino]

Missing the /data mount, not optional if you don’t change defaults../config is not optional..
Remove the whole health check for now.. pretty sure the indentation there is wrong for the interval/retries etc

[amrdeus2]

Thanks, appreciate all the help, but still not working.
Whenever I get indentation errors I just paste it in ChatGPT and that fixes that easy issue.
I've deleted the health checks and still get network related errors.
I've also noticed that whenever I delete all other containers in the stack and just leave transmission, it will deploy without any issues and create a stack network. So for instance, I originally manually created a network called "Arr" then quoted it in my code above. However when I deploy a stack it creates a network called "Stackname_Arr". So I think creating my own network isn't the issue either.

It sounds like its a port issue (from the errors and chatGPT), but I don't have a clue what the solution is.

[amrdeus2]

The following code gives me the error "failed to deploy a stack: Network arr_default Creating Network arr_default Created Container transmission-openvpn Creating Container transmission-openvpn Created Container sonarr Creating Container radarr Creating Container prowlarr Creating Error response from daemon: conflicting options: port publishing and the container type network mode"

`version: '3.3'

networks:
Arr:
external: true

services:
transmission-openvpn:
container_name: transmission-openvpn
cap_add:
- NET_ADMIN
volumes:
- /volume1/docker/ipvanish/:/etc/openvpn/custom/
- /media/USB/downloads:/downloads #optional
- /media/USB/data/transmission:/config #optional?
environment:
- OPENVPN_PROVIDER=custom
- OPENVPN_CONFIG=bulletVPN-AU-Sydney_1-syd01-tcp
- OPENVPN_USERNAME=XXXXX
- OPENVPN_PASSWORD=XXXXX
- LOCAL_NETWORK=192.168.0.0/16
- OPENVPN_OPTS=--inactive 3600 --ping 10 --ping-exit 60
logging:
driver: json-file
options:
max-size: 10m
ports:
- '9091:9091'
- '7878:7878' # port for radarr
restart: always
image: haugene/transmission-openvpn
networks:
- Arr

radarr:
image: lscr.io/linuxserver/radarr:latest
container_name: radarr
environment:
- puid=998
- pgid=100
- TZ=Australia/Sydney
volumes:
- /media/USB/data/radarr:/config
- /media/USB/data:/data #optional
ports:
- 7878:7878
restart: unless-stopped
network_mode: "service:transmission-openvpn" # Add to the transmission-openvpn Container Network
depends_on:
- transmission-openvpn # Set dependency on transmission-openvpn Container
networks:
- Arr
`

[amrdeus2]

Ok I think I got somewhere thanks to Chat GPT. I thought the documentation had it wrong but turns out I missed a step there:

        # You have to comment ports, they should be managed in transmission-openvpn section now.
#       ports:
#           - 9117:9117 

I didn't see the #s and had to scroll left to see them. This makes plenty of sense now. Apologies for the confusion!

The healthchecks are still giving me deploy errors, but I guess I don't need that...