2 minor fixes for kubeadm upgrade

Skipped prepulling etcd image if external etcd is used.

Fixes: kubernetes/kubeadm#1136

clean up redundant code for kubeadm join

kubeadm: Introduce config print init/join-defaults

Kubeadm graduate preflight phase

Remove temporary workaround for scheduler alpha config-loading

Delete repeated words and fix misspellings

Add api-machinery 'watch-consistency' e2e test

pass in stopCh to cloud provider Initialize method for custom controllers

register skeleton provider

debian-base: purge libsystemd0 to eliminate CVE false-positives

kubeadm: skip upgrade if manifest is not changed

fix reference to controlManagerExtraAargs

e2e refactor

Additionally, update the addon-manager to use kubectl v1.11.3.

docs/admin: add OWNERS file with cmd/kubeadm approvers

add more logging for azure disk diagnostics

kubectl: copy deployment util dependency into kubectl

Signed-off-by: yuexiao-wang <wang.yuexiao@zte.com.cn>

Remove error output from stdout to stderr

remove duplicate root check for kubeadm

delete node from aws if it is terminated

Fluentd: concatenate long logs

remove unneed if block for kubectl delete command

Fix error messages suggesting invalid flag