Set separate and clear resource requests/limits #54
Description
Currently the resources in the helm chart are a bit all over the place. For example the node driver registrar has the same resources as the container that run rclone in the daemonset. But the driver registrar does very little whereas the container that runs rclone on every node does a lot more. So their resources should be different.
Also the health check container needs very small and separate resources in the values file too.