Nodes running out of space

[carstenblt]

Description

My nodes just ran out of space because containerd kept all old images. Running

crictl rmi --prune

solved it. I am not sure yet about the best way to run this regularly but it might be worth adding?

[Silvest89]

Kubernetes automatically performs garbage collection
https://kubernetes.io/docs/concepts/architecture/garbage-collection/#containers-images

You can tune using the following kubelet args

• image-gc-high-threshold=80
• image-gc-low-threshold=75

[carstenblt]

Oh, I see, thank you!

So

k3s_exec_server_args = "--kubelet-arg=image-gc-high-threshold=50 --kubelet-arg=image-gc-low-threshold=40"

should do the trick?

[Silvest89]

Don't forget that you need to do the same for agent_args

And using those args wont work for existing nodes, and they won't persist across reboots.
I have opened a PR that allows to set kubelet-arg in the config.yaml instead

#1044

[maggie44]

Aren’t those args the default? k3s-io/k3s#813