hello, is the 'problem' with leftover containerd processes addressed somewhere? since rke2 updated to containerd 2.0, if a static pod is restarted (like kube-proxy), its former container-shim process is kept running. e.g., processes with pids: 1883573, 1907266, 3046680 are leftovers after control plane update 1.32.4 to 1.32.5 and later to 1.33.3

Hmm, I've not noticed that. Would you mind opening an issue?

shall I convert this to the issue or just create a new one? https://github.com/rancher/rke2/discussions/8331

Start a new issue if you wouldn't mind.

done

we’re upgrading to runc 1.3 and containerd 2.1 for next month, I’m curious if that’ll change anything

will gladly test, thank you