Scaleway has a great managed kubernetes offering. They host the control plane and all you pay are the instances starting at 30-something € per instance-month.
sixhobbits 18 minutes ago [-]
I've been running k3 on a single bigger hetzner dedicated machine to power https://demo.opensuite.online, a work-in-progress open source replacement for Google Workspace, a distribution of what the Dutch government is trying to switch to but they seemingly don't care about UX at all so I wanted to try build on their foundations (which in turn are basically just nextcloud, element, and some others).
Anyway can definitely recommend, it feels like a real cluster but it's all on a single piece of hardware which makes snapshots etc super simple.
pstuart 15 minutes ago [-]
> seemingly don't care about UX
That seems to be the case on every project with the folks that sign the checks or contracts for integrating other vendors.
I think Jobs was the rare exception there.
rsyring 33 minutes ago [-]
My experience with Hetzner VM pricing has turned me off of considering them for anything else. Maybe the experience in other world regions is not the same but this is what happened to us:
I started to use Hetzner's VM options in the Virginia, US datacenter. They had a great performance to cost ratio when I compared them to Linode/Akamai and Digital Ocean about a year ago. At that time, 4vcpu / 8GB server cost $24.99 per month.
Now, to get a VM like it, it's quoted $73.49 a month. Almost a 3x change.
I understand data center economics are changing due to LLM usage. But the Hetzner jump has not been mirrored by their competitors. I'm disappointed I made the switch from Linode. We still have some older VMs there and we'll likely continue to use them instead of Hetzner moving forward.
webology 23 minutes ago [-]
This has been my experience too. I migrated a few dozen projects from DO to Hetzner, and my monthly bill went from about $80 to around $162. The only thing that changed was Hetzner's rate hikes.
Now I'm in a weird spot. I could downgrade some servers, but changing them would immediately move me onto the new pricing. So I either pay more for less hardware, or pay a lot, lot more if I ever need to scale back up.
fabian2k 20 minutes ago [-]
They're the first, I strongly doubt they'll be the last to adjust prices that much. The big cloud providers might have enough margins to eat the cost increases, especially since bare VMs are only a small part of their offer. But I don't see how the smaller, more VM-focused providers will be able to absorb the high hardware prices.
Havoc 13 minutes ago [-]
You may have just caught the wrong end of it timing wise. They used to be super stable and cheap but being a low cost provider they have no real buffer to absorb even minor hikes let alone the shitshow we saw in ram/ssd.
throw1234567891 26 minutes ago [-]
Yeah, RAM shortage has caught up with them. Send a thank you note to Anthropic and OpenAI.
2001zhaozhao 23 minutes ago [-]
> Cheap
> Hetzner cloud
> published in December 2025
which means that it's twice as expensive now as it was when the article was published
stefan_ 8 minutes ago [-]
So are computers
Havoc 15 minutes ago [-]
Went a slightly different angle - Rasp 4s aren't impressing anyone these days, but 3 of them make for a fine HA controlplane. Once you've got a stable HA controlplane its much easier to connect whatever hardware you have lying around
(Just don't use SD cards...use M10 optanes). Also...Talos not k3s.
stubbi 26 minutes ago [-]
I developed a Karpenter provider for Hetzner recently. Might interesting for you to integrate.
Terraform has been such a delight in the age of LLMs.
embedding-shape 30 minutes ago [-]
"Cheap" in terms of money, ~30K in terms of lines of code that you eventually gonna need to bite the apple and read through and understand, 10K of which is HCL.
End cost seems to be ~60 EUR/month for a cluster of machines, which isn't too bad. Still, I suppose once you're even dealing with such small amounts, even thinking about distributed architecture or similar things feels a bit too early.
timwis 29 minutes ago [-]
For those interested in this I also recommend autobase, a set of ansible roles for a highly available postgres cluster on hetzner or other clouds: https://github.com/autobase-tech/autobase
15 minutes ago [-]
bakies 37 minutes ago [-]
You really put your admin credentials in a blog post. Lucky I can't seem to hit that IP.
Rendered at 19:35:30 GMT+0000 (Coordinated Universal Time) with Vercel.
Anyway can definitely recommend, it feels like a real cluster but it's all on a single piece of hardware which makes snapshots etc super simple.
That seems to be the case on every project with the folks that sign the checks or contracts for integrating other vendors.
I think Jobs was the rare exception there.
I started to use Hetzner's VM options in the Virginia, US datacenter. They had a great performance to cost ratio when I compared them to Linode/Akamai and Digital Ocean about a year ago. At that time, 4vcpu / 8GB server cost $24.99 per month.
Now, to get a VM like it, it's quoted $73.49 a month. Almost a 3x change.
I understand data center economics are changing due to LLM usage. But the Hetzner jump has not been mirrored by their competitors. I'm disappointed I made the switch from Linode. We still have some older VMs there and we'll likely continue to use them instead of Hetzner moving forward.
Now I'm in a weird spot. I could downgrade some servers, but changing them would immediately move me onto the new pricing. So I either pay more for less hardware, or pay a lot, lot more if I ever need to scale back up.
> Hetzner cloud
> published in December 2025
which means that it's twice as expensive now as it was when the article was published
(Just don't use SD cards...use M10 optanes). Also...Talos not k3s.
https://github.com/paperclipinc/karpenter-provider-hetzner
End cost seems to be ~60 EUR/month for a cluster of machines, which isn't too bad. Still, I suppose once you're even dealing with such small amounts, even thinking about distributed architecture or similar things feels a bit too early.