Replies: 3 comments 4 replies
-
Pretty sure no one has, and even as a NixOS user myself, I don't really see the cost-benefit analysis too positive for such a project. MicroOS works well enough for the current approach and we can use it pretty much as-is. While NixOS offers much more advanced customization-options and tooling, but comes with a far higher learning-curve (not sure if anyone else on the team used nix before). Whats your motivation behind this? Do you want to install NixOS via terraform (already possible with various projects, sometimes a bit hacky). Or would you like to run k3s on NixOS? There's already a service module upstream. Or is it about the hetzner-specific features like CCM and so on? |
Beta Was this translation helpful? Give feedback.
-
Hello @otavio, I agree with @phaer, as much as NixOS can be nice. MicroOS is ideal for the job, it's container optimized, it's transactional in nature, it's has a low attack surface and comes hardened out of the box, it's updated daily, as no major version updates, it just updates all the time infinitely (it's based on opensuse tumbleweed), and handles reboot-need signaling well to let Kubernetes know that a reboot is needed after an upgrade. So it's really ideal! |
Beta Was this translation helpful? Give feedback.
-
That's not in our plans for now, we are moving to the more stable LeapMicro. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
I am looking for alternatives to host our cluster and fund this project; it is terrific! I want to ask if someone evaluated the use of NixOS as host distribution?
Beta Was this translation helpful? Give feedback.
All reactions