16
Nuked and back again - a tale of losing and recovering a bare-metal kubernetes cluster (mostly)
(blog.keyboardvagabond.com)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 06 Dec 2025
16 points (100.0% liked)
Programming
23729 readers
164 users here now
Welcome to the main community in programming.dev! Feel free to post anything relating to programming here!
Cross posting is strongly encouraged in the instance. If you feel your post or another person's post makes sense in another community cross post into it.
Hope you enjoy the instance!
Rules
Rules
- Follow the programming.dev instance rules
- Keep content related to programming in some way
- If you're posting long videos try to add in some form of tldr for those who don't want to watch videos
Wormhole
Follow the wormhole through a path of communities !webdev@programming.dev
founded 2 years ago
MODERATORS
Noooo this is possibly the worst case I've seen of why the fuck did you complicate things this much?? I blame Claude/Cursor mostly for this one, it seems they steered a lot of the design decisions. Seriously though, if you think you need a bare bones kubernetes setup on 3 small hetzner vpses, only two of which being worker nodes, with a grand total 16gb of RAM, to support fedi instances with around about a single user... Stop. Just stop it already. I obviously feel bad that everything crashed down on them, but if they had taken the time to actually talk to other small instance maintainers, they could've spent more time doing proper backup and recovery than messing around with, and I shit you not, I counted 11 unique tools. Nobody needs 11 monitoring, containerization, volume management, dns tools, fucking cloudflare?? To run a single user instance.
the instances were an excuse for me to set up a cluster. I definitely know none of this isn't needed. I wanted to get familiar with things that I do related to work, but never really get to get my hands dirty with. I did all of this because I wanted to learn.
I suggest framing the blog post this way. The way it reads now is like you're actually suggesting this to other small instance admins. I think it's OK to play around with new tools, but the article is not self aware in this respect.
EDIT: it's also comes off like, oh no one silly keystroke nuked my setup, but from an outside perspective, of course it did, it was a house of cards. I think a better framing would be that you dug a large hole for yourself by trying to force enterprise level complexity into a one-person project, and through sheer perseverance it somehow is still alive, but maybe some different architectural decisions early on could have avoided this.