I created a GitHub template and set of workflows for creating Pi K3S clusters in my home lab, which I've recently tidied up, documented, and decided to publish.
My project aims to automate the configuration of bare metal K3S clusters in a simple, repeatable fashion, and get you from first commit to first pod as quickly as possible. It should be* as simple as: gather your hardware, clone the GitHub template, push your cluster configuration, wait for the custom OS image to build, copy it to a storage device, and boot.
I have found that it is possible (just!) to get a multi-node cluster up and running from source in under 30 minutes. Required configuration is minimal - most settings have defaults or are optional. All configuration options are documented.
Scooby doesn't use anything fancy like Ansible for automation; most of the work is performed by shell scripts so it should be quite easily accessible for newcomers.
I hope that by aggregating everything I've learned into a set of easily repeatable steps, I can save others beginning their bare metal journey some time and headaches.
Thanks for reading. If you do try it out, I would love some feedback :)
* there are a few small gotchas but don't worry, it's all in the README!
You do know that there is no Yammer or Skype in London anymore? (we did have a fantastic culture though, so great it has been recognised. source: ex-Yammer.)
My project aims to automate the configuration of bare metal K3S clusters in a simple, repeatable fashion, and get you from first commit to first pod as quickly as possible. It should be* as simple as: gather your hardware, clone the GitHub template, push your cluster configuration, wait for the custom OS image to build, copy it to a storage device, and boot.
I have found that it is possible (just!) to get a multi-node cluster up and running from source in under 30 minutes. Required configuration is minimal - most settings have defaults or are optional. All configuration options are documented.
Scooby doesn't use anything fancy like Ansible for automation; most of the work is performed by shell scripts so it should be quite easily accessible for newcomers.
I hope that by aggregating everything I've learned into a set of easily repeatable steps, I can save others beginning their bare metal journey some time and headaches.
Thanks for reading. If you do try it out, I would love some feedback :)
* there are a few small gotchas but don't worry, it's all in the README!