WebUI inaccessible

[4xgkg]

I've been using the container :latest for a few years now.

I used to regularly pull the latest.

A couple of weeks ago, the web UI stopped responding.

I have LOCAL_NETWORK set correctly to my home LAN CIDR, as is also described in https://haugene.github.io/docker-transmission-openvpn/vpn-networking/ and this used to work fine for a long time.

Here is the current docker-compose.yml (after many tweaks to try to fix this):

version: '3.3'
services:
 transmission:
  image: haugene/transmission-openvpn:4.2
  container_name: transmission
  cap_add:
    - NET_ADMIN
  privileged: true
  devices:
    - /dev/net/tun
  restart: unless-stopped
  ports:
    - "9091:9091"
    - "8118:8118"
  dns:
    - 8.8.8.8
    - 8.8.4.4
  volumes:
    - /etc/localtime:/etc/localtime:ro
    - ${HOME}/Downloads/trans/:/data
    # This is a new one, from following the latest docs
    - ${HOME}/transmission/config:/config
  environment:
    - CREATE_TUN_DEVICE=true
    - OPENVPN_PROVIDER=<reducted>
    - OPENVPN_CONFIG=<reducted>
    - OPENVPN_USERNAME=<reducted>
    - OPENVPN_PASSWORD=<reduced>
    - OPENVPN_OPTS=--inactive 3600 --ping 10 --ping-exit 60
    - WEBPROXY_ENABLED=true
    - WEBPROXY_PORT=8118
    - LOCAL_NETWORK=192.168.0.0/24
    - PUID=1000
    - PGID=1000

Executing netstat and the logs show that the VPN comes up fine, even Transmission finds files in the watch directory and pulls them, but I can't reach the WebUI.

root@6580534535a8:/# netstat -tlpn
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address           Foreign Address         State       PID/Program name
tcp        0      0 127.0.0.11:39635        0.0.0.0:*               LISTEN      -
tcp        0      0 172.21.0.2:8118         0.0.0.0:*               LISTEN      104/privoxy
tcp        0      0 10.8.8.7:57707          0.0.0.0:*               LISTEN      105/transmission-da
tcp6       0      0 :::57707                :::*                    LISTEN      105/transmission-da

Privoxy itself is accessible on port 8118, and lets me tunnel my HTTP traffic through the VPN:

❯ curl -x http://transmission-host-ip:8118/ ifconfig.co/json
{
...
  "country": "Singapore",
  "user_agent": {
    "product": "curl",
    "version": "7.79.1",
    "raw_value": "curl/7.79.1"
  }
}

(Singapore is the VPN location)

But trying to reach port 9091 (which I also see is not listned to on the container), gets Connection Refused:

❯ curl -v http://transmission-host-ip:9091/
*   Trying transmission-host-ip:9091...
* connect to transmission-host-ip port 9091 failed: Connection refused
* Failed to connect to nuc port 9091 after 126 ms: Connection refused
* Closing connection 0
curl: (7) Failed to connect to nuc port 9091 after 126 ms: Connection refused

What am I missing?

[4xgkg]

More digging:

I found in /data/transmission-home/settings.json:

settings.json:    "rpc-bind-address": "9.0.0.0",

And in transmission.log:

[2022-11-11 14:27:24.533] RPC Server Unable to bind to 9.0.0.0:9091 after 10 attempts, giving up (rpc-server.c:845)

This makes sense since this isn't a valid address. I'm beginning to suspect a bug in one of the setup scripts picking the port number of 9091 and doing something weird with it.

EDIT: I edited the file to set rpc-bind-address to 0.0.0.0 but a docker-compose restart reset it back to 9.0.0.0, so something is overwriting it during startup.

[pkishino]

stop the container, set it to 0.0.0.0, then start the container.
you cannot edit the settings.json file while the container runs

[4xgkg]

stop the container, set it to 0.0.0.0, then start the container. you cannot edit the settings.json file while the container runs

That worked!
I also moved away settings.json and it still works.
So somehow, this file got munged and it completely destroyed my experience.

Thanks!

[deathblade666]

I'm also having this issue, however my rpc_bind_address is 0.0.0.0 but my bind_address_ipv4 is not (not sure if that setting matters here) i have tried to change it but it always reverts regardless of container status (running, stopped, killed (with compose down), etc)

[deathblade666]

oh i see the bind_address_ipv4 is for the tunnel. then not sure why its not loading the web UI for me just get a timeout error

UPDATE: When I run a curl against it I get a 301 error

UPDATE2: found I had to access it from <host_ip>:9091/transmission/web/ can't access directly from <host_ip>:9091

[pkishino]

that is as expected per default url scheme for transmission
, you can change this by changing TRANSMISSION_RPC_URL as per https://github.com/transmission/transmission/blob/main/docs/Editing-Configuration-Files.md#options

[KrazyKaiser]

I'm still having trouble accessing the web ui even when i go to <host_ip>:9091/transmission/ or <host_ip>:9091/transmission/web, any advice? Tranmission is up and running properly, sonarr can use it to download things. I just can't go to it myself.