NDM should watch for ConfigMap changes

[kmova]

NDM DaemonSet pods use the ConfigMap to determine what disks are discovered and added as disk resources. This configuration is read at the startup by the DaemonSet Pods.

What is the expected behavior? It will be good to have the pods watch for the ConfigMap changes and dynamically (at runtime) change the logic for types of disks discovered.

This also requires the NDM to handle cases, where some of the disks that were discovered with previous config to be deleted, if the new config restricts them.

[davidkarlsen]

When managing the install with helm, this trick can be used: https://github.com/helm/charts/blob/master/stable/cerebro/templates/deployment.yaml#L29 as a change in configmap will trigger a change in annotation which again triggers a rollout.