Currently we are completing kubeadm init while the scheduler and controller manager are likely still not yet up. In some cases if they will fail, they won't ever come up. Instead wait until each pod enters running state before exiting kubeadm.