Description of problem:
See the following details.

Version-Release number of the following components:
openshift-ansible-3.7.0-0.127.0.git.0.b9941e4.el7.noarch

How reproducible:
Always

Steps to Reproduce:
1. prepare inventory host file, enable cri-o, and define a standalone nfs host
2. trigger installation
3.

Actual results:
Installation failed due to installer is trying run cri-o on a nfs host.

fatal: [ec2-54-86-209-24.compute-1.amazonaws.com]: FAILED! => {
    "changed": true, 
    "cmd": "lsmod | grep overlay", 
    "delta": "0:00:00.006780", 
    "end": "2017-09-22 04:16:35.417998", 
    "failed": true, 
    "rc": 1, 
    "start": "2017-09-22 04:16:35.411218"
}

MSG:

non-zero return code

<--snip-->

fatal: [ec2-54-86-209-24.compute-1.amazonaws.com]: FAILED! => {
    "changed": false, 
    "failed": true
}

MSG:

Unable to start service cri-o: Job for cri-o.service failed because the control process exited with error code. See "systemctl status cri-o.service" and "journalctl -xe" for details.
<--snip-->

Failure summary:


  1. Hosts:    ec2-54-86-209-24.compute-1.amazonaws.com
     Play:     Set openshift_version for all hosts
     Task:     Start the CRI-O service
     Message:  Unable to start service cri-o: Job for cri-o.service failed because the control process exited with error code. See "systemctl status cri-o.service" and "journalctl -xe" for details.


Expected results:
Installer should never install cri-o service on a nfs host.


Additional info:
docker rpm never is installed on a nfs host, cri-o service also should never installed on a nfs host, should only installed on a node host.