Modifications to kubeVersion or targets of lookup calls should result in automatic Helm redeployment

[aiyengar2]

Currently, the k3s-io/helm-controller does not support auto-upgrading helm charts on seeing changes to the underlying cluster state that would result in a requirement to re-evaluate the Helm chart's contents.

This needs to be patched for k3s-io/helm-controller#141

[aiyengar2]

Workaround

Simply delete the helm-install-<release-name> Job tied to the HelmChart tied to the ProjectHelmChart that needs the upgrade; this will be located in the same namespace that the operator is deployed into (i.e. cattle-helm-system for Helm Project Operator, cattle-monitoring-system for Prometheus Federator); once deleted, Helm Project Operator should automatically recreate it, which will trigger a Helm upgrade that will pick up the changes based on the new cluster state.