fix scaling in kubeadm etcd mode (#6822)

'ansible.vars.hostvars.HostVarsVars object' has no attribute 'kubeadm_upload_cert'

kubeadm_upload_cert will never be found as a hostvar for the first
master since the task is executed for a worker.

Fix by executing the upload task for the first master and register
the needed key. After that, workers can read hostvars for the master

Var kubeadm_etcd_refresh_cert_key removed since it no longer has
any use.

roles/kubernetes/kubeadm/defaults/main.yml

@@ -11,8 +11,5 @@ kube_override_hostname: >-
   {{ inventory_hostname }}
   {%- endif -%}
 
-# Requests a fresh upload of certificates from first master
-kubeadm_etcd_refresh_cert_key: true
-
 # Experimental kubeadm etcd deployment mode. Available only for new deployment
 etcd_kubeadm_enabled: false

roles/kubernetes/kubeadm/tasks/kubeadm_etcd_node.yml

@@ -1,18 +1,7 @@
 ---
-- name: Refresh certificates so they are fresh and not expired
-  command: >-
-    {{ bin_dir }}/kubeadm init phase
-    --config {{ kube_config_dir }}/kubeadm-config.yaml
-    upload-certs
-    --upload-certs
-  register: kubeadm_upload_cert
-  delegate_to: "{{ groups['kube-master'][0] }}"
-  when: kubeadm_etcd_refresh_cert_key
-  run_once: yes
-
 - name: Parse certificate key if not set
   set_fact:
-    kubeadm_certificate_key: "{{ hostvars[groups['kube-master'][0]]['kubeadm_upload_cert'].stdout_lines[-1] | trim }}"
+    kubeadm_certificate_key: "{{ hostvars[groups['kube-master'][0]]['kubeadm_certificate_key'] }}"
   when: kubeadm_certificate_key is undefined
 
 - name: Pull control plane certs down