Hey all,

I deployed tailscale to my local network to a lab rpi running pontainer - no surprises here....

The problem is, that the tailscale daemon running inside the container seems to b given the TF_ROUTES in my compose file. Nope.... It exposed nothing until I connected to the container and ran `tailscale set --advertise-routes= providing the authkey as an argument.

version: "3.3"
services:
  tailscale:
      privileged: true
      hostname: tailscale
      network_mode: host
      container_name: tailscale
      image: tailscale/tailscale:latest
      volumes:
          - "/mnt/tailscale/var:/var/lib"
          - "/dev/net/tun:/dev/net/tun"
      cap_add:
        - NET_ADMIN
        - NET_RAW
      environment:
        - PUID=1000
        - PGID=1000
        - TS_USERSPACE=true
        - TS_AUTHKEY=aaa
        - TS_ROUTES=192.168.1.0/24
        - TS_HOSTNAME=tailscale
        - TS_STATE_DIR=/var/lib/tailscale
      command: tailscaled
      restart: unless-stopped

The problem is, that the tailscale daemon running inside the container seems to be ignoring the TS vars provided - this means, the device did not authorized, unless I exec'ed to the container and run the tailscale command manually, providing the authkey as an argument.

The same with the routes - I was expecting the device would expose 192.168.1.0/24 given the the TF_ROUTES in my compose file. Nope.... It exposed nothing until I connected to the container and ran `tailscale set --advertise-routes=192.168.1.0/24` - after doing so, the route works as expected and was available for an approval in the web ui.

Any idea why? I thought I am providing the env variables correctly and these will be used by tailscaled but these seem to be ignored..

Thanks for any help!