The Fabric8 scale(...) method IIRC used to wait for readiness of the new pods but doesn't do so anymore. I think we fixed that for Kafka pods by adding additional readiness pod. But when scaling up Zookeeper, we currently start the new pod and even before it is ready we start rolling the old pods. That can for sure have some impact on availability because the new pod will not be ready while we take down the first old pod.

This PR adds additional readiness check for the new pod to make sure it is ready before we move to the rolling update of the old pods.