Description of problem:
Run redeploy-certificates.yml playbook against an ocp-3.10 cluster, playbook may fail as below:

PLAY [Restart nodes] *******************************************************************************
...

TASK [Restart docker] *******************************************************************************************************************************************************
changed: [ec2-52-90-247-129.compute-1.amazonaws.com] => {"attempts": 1, "changed": true, "failed": false, "name": "docker", "state": "started", ...


TASK [Wait for master API to come back online] ******************************************************************************************************************************
ok: [ec2-52-90-247-129.compute-1.amazonaws.com] => {"changed": false, "elapsed": 25, "failed": false, "path": null, "port": 8443, "search_regex": null, "state": "started"}

TASK [restart node] *********************************************************************************************************************************************************
changed: [ec2-52-90-247-129.compute-1.amazonaws.com] => {"changed": true, "failed": false, "name": "atomic-openshift-node", "state": "started", "status": ...

TASK [Wait for node to be ready] ********************************************************************************************************************************************
fatal: [ec2-52-90-247-129.compute-1.amazonaws.com]: FAILED! => {"failed": true, "msg": "The conditional check 'node_output.results.returncode == 0 and node_output.results.results[0].status.conditions | selectattr('type', 'match', '^Ready$') | map(attribute='status') | join | bool == True' failed. The error was: error while evaluating conditional (node_output.results.returncode == 0 and node_output.results.results[0].status.conditions | selectattr('type', 'match', '^Ready$') | map(attribute='status') | join | bool == True): 'dict object' has no attribute 'results'"}
	to retry, use: --limit @/usr/share/ansible/openshift-ansible/playbooks/redeploy-certificates.retry



Sometimes master-api service still not available when running "TASK [Wait for node to be ready] ", since docker restart will also trigger master-api/controllers service restart, so maybe we need "verify API server" task before restart node.


Version-Release number of the following components:
openshift-ansible-3.10.0-0.58.0.git.0.d8f6377.el7.noarch


How reproducible:
50% (3 failures in 6 attempts)

Steps to Reproduce:
1.Run openshift cert redeploy playbook
ansible-playbook -i host/310 -vvv /usr/share/ansible/openshift-ansible/playbooks/redeploy-certificates.yml


Actual results:
Playbook fails as in Description, then log into master, run "/usr/bin/oc get node ip-172-18-11-10.ec2.internal -o json -n default", it could work.


Expected results:


Additional info:
Ansible inventory file and full log with "-vvv" could be found in attachment