Go to file
2024-05-20 08:53:46 -04:00
.github/workflows Fixes the format of the run on the date/time workflow. 2022-02-12 17:18:12 -05:00
.gitattributes CHANGE: Add a default index.html file that enables autoreconnect and resizing. 2024-04-04 11:04:18 -04:00
build_and_push.sh Simplified the Dockerfile further and also fixed the build and push script to properly push the tagged builds. 2022-02-07 20:21:28 -05:00
docker-compose.build.yml Add docker-compose.build.yml 2023-12-04 09:33:32 -06:00
docker-compose.yml fix: update docs and example compose 2024-04-04 10:12:49 -04:00
Dockerfile CHANGE: Add icons for Prusaslicer to replace the noVNC icons to better indicate tabs. 2024-05-19 01:09:58 -04:00
entrypoint.sh fix: remove debug 2024-04-04 10:17:03 -04:00
get_latest_prusaslicer_release.sh feat: test moving to GTK3. Looks much more modern 2024-04-03 16:28:42 -04:00
README.md Merge pull request #16 from tiwiz/patch-1 2024-05-19 01:13:27 -04:00
supervisord.conf fix: update to latest version and associated deps. and setup VGLRUN 2024-04-03 13:09:07 -04:00
tag_latest_prusaslicer.sh * Simplified the Dockerfile further. 2022-02-05 13:41:50 -05:00
vncresize.html CHANGE: Add a default index.html file that enables autoreconnect and resizing. 2024-04-04 11:04:18 -04:00

Prusaslicer noVNC Docker Container

Overview

This is a super basic noVNC build using supervisor to serve Prusaslicer in your favorite web browser. This was primarily built for users using the popular unraid NAS software, to allow them to quickly hop in a browser, slice, and upload their favorite 3D prints.

A lot of this was branched off of dmagyar's awesome prusaslicer-vnc-docker project, but I found it to be a bit complex for my needs and thought this approach would simplify things a lot.

How to use

In unraid

If you're using unraid, open your Docker page and under Template repositories, add https://github.com/helfrichmichael/unraid-templates and save it. You should then be able to Add Container for prusaslicer-novnc. For unraid, the template will default to 6080 for the noVNC web instance.

Outside of unraid

Docker

To run this image, you can run the following command: docker run --detach --volume=prusaslicer-novnc-data:/configs/ --volume=prusaslicer-novnc-prints:/prints/ -p 8080:8080 -e SSL_CERT_FILE="/etc/ssl/certs/ca-certificates.crt" --name=prusaslicer-novnc mikeah/prusaslicer-novnc

This will bind /configs/ in the container to a local volume on my machine named prusaslicer-novnc-data. Additionally it will bind /prints/ in the container to superslicer-novnc-prints locally on my machine, it will bind port 8080 to 8080, and finally, it will provide an environment variable to keep Prusaslicer happy by providing an SSL_CERT_FILE.

Docker Compose

To use the pre-built image, simply clone this repository or copy docker-compose.yml and run docker compose up -d.

To build a new image, clone this repository and run docker compose up -f docker-compose.build.yml --build -d

Using a VNC Viewer

To use a VNC viewer with the container, the default port for TurobVNC is 5900. You can add this port by adding -p 5900:5900 to your command to start the container to open this port for access. See note below about ports related to VNC_PORT environment variable.

GPU Acceleration/Passthrough

Like other Docker containers, you can pass your Nvidia GPU into the container using the NVIDIA_VISIBLE_DEVICES and NVIDIA_DRIVER_CAPABILITIES envs. You can define these using the value of all or by providing more narrow and specific values. This has only been tested on Nvidia GPUs.

In unraid you can set these values during set up, additionally, please add the "Extra Parameters" with --runtime=nvidia to ensure the GPU passes through. For containers outside of unraid, you can set this by adding the following params or similar -e NVIDIA_DRIVER_CAPABILITIES="all" NVIDIA_VISIBLE_DEVICES="all". If using Docker Compose, uncomment the enviroment variables in the relevant docker-compose.yaml file.

In addition to the information above, to enable Hardware 3D acceleration (which helps with visualizing complex models and sliced layers), you must set an environment variable. You can do this by either adding -e ENABLEHWGPU=true to the docker run command or including - ENABLEHWGPU=true in your Docker Compose configuration.

Once enabled and started you can verify the GPU is being used by running nvidia-smi -l on the HOST machine and you should see /slic3r/slic3r-dist/bin/prusa-slicer as process using the GPU.

+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.161.07             Driver Version: 535.161.07   CUDA Version: 12.2     |
|-----------------------------------------+----------------------+----------------------+

.. removed for brevity .. 

+---------------------------------------------------------------------------------------+
| Processes:                                                                            |
|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
|        ID   ID                                                             Usage      |
|=======================================================================================|
|    0   N/A  N/A   4129827      G   /slic3r/slic3r-dist/bin/prusa-slicer        262MiB |
+---------------------------------------------------------------------------------------+

*some information above was edited for privacy

The GL Version on the System Information screen inside the slicer should also show, the GPU model and driver version

Other Environment Variables

Below are the default values for various environment variables:

  • DISPLAY=:0: Sets the DISPLAY variable (usually left as 0).
  • SUPD_LOGLEVEL=INFO: Specifies the log level for supervisord. Set to TRACE to see output for various commands helps if you are debugging something. See superviosrd manual for possible levels.
  • ENABLEHWGPU=: Enables HW 3D acceleration. Default is false to maintain backward compatability.
  • VGL_DISPLAY=egl: Advanced setting to target specific cards if you have multiple GPUs
  • NOVNC_PORT=8080: Sets the port for the noVNC HTML5/web interface.
  • VNC_RESOLUTION=1280x800: Defines the resolution of the VNC server.
  • VNC_PASSWORD=: Defaults to no VNC password, but you can add one here.
  • VNC_PORT=5900: Defines the port for the VNC server, allowing direct connections using a VNC client. Note that the DISPLAY number is added to the port number (e.g., if your display is :1, the VNC port accepting connections will be 5901).

Prusaslicer

TruboVNC

VirtualGL

Supervisor

GitHub Source

Docker

Buy Me A Coffee