Currently, the health check after the deployment checks that all the nodes returned by kubectl are ready; however, the worker might not be attached yet to the cluster or may have issues.
Here, the UI said it's ready, but when I ssh to the master, I found that the master was the only one there, and the worker wasn't attached yet.
