Octoprint with socat for remote klipper/printers

This is a Dockerfile to set up OctoPrint alongside socat to allow remote connections to 3d printers or klipper instalattions.

This solution is a compilations various ideas found on related topics, but at its core it uses the linixserver.io's alpine base image to take advantage of the s6 overlaying and a image format well known.

Why?

My needs are very simple, I want to:

Run octoprint in a hardware more powerful than a Rpi for a snappier experience;
Set UID and GUID of the user running octoprint;
Keep installed octoprint plugins even when container is destroyed;
Connect to a remote 3Dprinter, ESP3D or klipper

How?

I achieved this solution by compiling the information in the Octoprint's docker GitHub and S6 documentations

Some information that I consider useful about this project:

UID and GUID of user running Octoprint can be overridden like on any other linuxserver.io image;
The said user has sudo permissions allowing to execute system commands like restart Octoprint itself or the socat like: sudo s6-svc -r /var/run/s6/services/octoprint or sudo s6-svc -r /var/run/s6/services/socat
All Octoprint's configuration files and installed plugins are present at /config (equivalent to the .octoprint)

What's Next?

Use GitHubs actions to build and publish to DockerHub
Allow multi ARCH (not my top priority as the main reason of this project is performance, so amd64)

Usage

Usual workflow with docker-compose build and docker-compose up cpmmands Or import on Portainer

Create a socat service on klipper host. I included the socat_klipper.service file as example: Copy to /etc/systemd/system/ and do sudo systemctl enable socat_klipper.service and after sudo systemctl start socat_klipper.service

And you're good to go.

Environment options:

All Octoprint variables are available and on top of that a few others have been added:

SOCAT_USER=abc User to have permissions to access this socat interface

SOCAT_PRINTER_HOST="localhost" SOCAT_PRINTER_PORT="5555" Where socat should connect to - will be used as tcp://localhost5:5555