Add check/warning for pre-existing ClusterRole in Kubernetes example scripts
diverdane opened this issue · comments
Is your feature request related to a problem? Please describe.
If you are using the authn-k8s demo example in examples/kubernetes interatively,
and you do the following sequence:
- Run the kubernetes authn-k8s demo
- Delete the demo namespace (e.g.
kubectl delete namespace app-test) - Instead of deleting the Helm release (e.g. with
helm delete conjur-oss),
you simply delete the Conjur OSS namespace (e.g. withkubectl delete namespace conjur-oss)
(This isn't the documented way to clean up, but hey, it happens.) - Re-run the demo (e.g. with
./start)
Then you will see errors during Helm upgrade saying that you already have a
ClusterRole object.
Describe the solution you would like
In the examples/kubernetes/2_helm_install_or_upgrade_conjur.sh script, add
a check for a pre-existing ClusterRole, and if there is one, perhaps
display a warning and set the chart value to not create the cluster RBAC.