This message was deleted.
# rke2
a
This message was deleted.
p
Screenshoot with the logs
By the way, if I spin up an rke2 cluster with the same config locally in the Node and then import it on Rancher. It works like a charm. 😮
c
if you disable the built-in ccm, you need to replace it with something else before the cluster agent pod will be able to run and phone home to rancher.
If you’re using the UI, the recommended way to do that would be to pass an additional manifest that installs the CCM you want via HelmChart, with bootstrap set to true in the HelmChart resource. https://docs.rke2.io/helm#using-the-helm-crd
p
@creamy-pencil-82913 Thanks for your answer, I tested again after your guidance with the following configuration as an additional manifest but still not working. Am I missing something? apiVersion: helm.cattle.io/v1 kind: HelmChart metadata: name: hcloud-cloud-controller-manager namespace: kube-system spec: version: v1.18.0 repo: https://charts.hetzner.cloud chart: hcloud-cloud-controller-manager targetNamespace: kube-system bootstrap: true valuesContent: |- env: HCLOUD_NETWORK: HCLOUD_TOKEN:
c
That looks good. What about it is not working? Does the helm chart fail to install? Does it install but not work?
You probably need to disable the default cloud controller manager if you’re going to deploy your own
p
I don't remember if i disabled the ccm, but the Cluster is still "stuck" and not registering. I will check again and come back with some feedback.