We provide configuration at deployment time via the
--set parameter which is supported for both
helm install and
helm upgrade . These override defaults in the
values.yaml file, which are then either mapped to a
Finally, those values are then injected as environment variables to the pods that require them.
This all works pretty well for us.