Welcome to the Linux Foundation Forum!

Lab 4.1 - cluster installation fails due to kuberhealthy

laurent.roffe
laurent.roffe Posts: 17
edited September 2022 in LFS268 Class Forum

Hi all,

The cluster installation seem not so straightforward on GCP, with terraform apply command

I had some issues with IAM accounts creation failing, some errors due to number of occurrence limitation of google API, all that was solved by retrying again several times the command tf apply (I also manually upgraded the k8s cluster version to 1.23 directly in the GCP console)

at the end there was still an error seemingly due to kuberhealthy, with a CRD ressource not found in the API version mentioned. here is the error :

module.jx.module.cluster.module.jx-health[0].helm_release.kuberhealthy: Creating...
╷
│ Error: failed to install CRD crds/khcheck.yaml: resource mapping not found for name: "khchecks.comcast.github.io" namespace: "" from "": no matches for kind "CustomResourceDefinition" in version "apiextensions.k8s.io/v1beta1"
│ ensure CRDs are installed first
│ 
│   with module.jx.module.cluster.module.jx-health[0].helm_release.kuberhealthy,
│   on .terraform/modules/jx.cluster.jx-health/main.tf line 1, in resource "helm_release" "kuberhealthy":
│    1: resource "helm_release" "kuberhealthy" {

I found that this issue seem to be logged at the link below (quite recently) and that fortunately it is possible to install the cluster without kuberhealthy :

there

Let's hope that this partial installation will not geopardize the rest of the training and that this error will be fixed soon

I would appreciate any advise or hint, to troubleshoot this error, since for now I lack too much knowledge about terraform and its integration with helm to be able to identify the root cause of this issue

Laurent

Comments

  • Same here, I cannot deploy as per the instructions due to kuberhealthy. Thank you for the workaround Laurent.

  • Yup, same for me, but updating the values with kuberhealthy = false (after deleting both local / github repos > re cloning > pushing back to origin) returned with the same issue.

Categories

Upcoming Training