Fail to connect port 8080

[Fisher-Wang]

My environment

• Ubuntu 22.04.4 LTS
• Nvidia Driver version: 560.35.03
• Nvidia Container version:

  libnvidia-container-dev/unknown 1.16.1-1 amd64
  libnvidia-container-tools/unknown,now 1.16.1-1 amd64 [installed,automatic]
  libnvidia-container1-dbg/unknown 1.16.1-1 amd64
  libnvidia-container1/unknown,now 1.16.1-1 amd64 [installed,automatic]
  nvidia-container-runtime/unknown 3.14.0-1 all
  nvidia-container-toolkit-base/unknown,now 1.16.1-1 amd64 [installed,automatic]
  nvidia-container-toolkit/unknown,now 1.16.1-1 amd64 [installed]
  

How to reproduce

1. git clone https://github.com/selkies-project/docker-nvidia-glx-desktop
   cd docker-nvidia-glx-desktop
   

2. uncomment network_mode: host in docker-compose.yml

3. sudo docker compose up -d

4. nc -zv localhost 8080

   Then I got the error nc: connect to localhost (127.0.0.1) port 8080 (tcp) failed: Connection refused

5. Open a brower and visit http://127.0.0.1:8080

   Then I got this
   

Could you help me run it correctly? Thank you very much

[ehfd]

Need logs inside the container located cat /tmp/*.log

[Fisher-Wang]

Hi, here is all my log files at /tmp/*.log

/tmp/dbus.log
/tmp/entrypoint.log
/tmp/selkies-gstreamer-entrypoint.log
/tmpsupervisord.log
/tmp/kasmvnc-entrypoint.log: empty
/tmp/nginx.log: empty
/tmp/pipewire.log: empty
/tmp/pipewire-pulse.log: empty
/tmp/wireplumber.log: empty

[PMohanJ]

As per logs, it seems like Xserver if failing. By default the DISPLAY is set to :20 and there's comment in compose file saying running on host network with :20 might lead to error https://github.com/selkies-project/docker-nvidia-glx-desktop/blob/main/docker-compose.yml#L28.
Could set the env var DISPLAY to a random value, for instance DISPLAY=77 and try again?

[Fisher-Wang]

I add - DISPLAY=:77 to the environment entry of docker-compose.yml, and restart the docker. Unfortunately it gives the same logs as above.

[Fisher-Wang]

And here is the content of /home/ubuntu/.local/share/xorg/Xorg.77.log: https://pastebin.ubuntu.com/p/4wZ9XGQkDy/, which may have some useful information

[PMohanJ]

I'm not sure of those Xorg logs. @ehfd might look at those when he gets time.

[ehfd]

This seems like a textbook case of having an X server already running on that GPU in the host.

Answer: #11