Change DNS policy for kubernetes components

According to code apiserver, scheduler, controller-manager, proxy don't
use resolution of objects they created. It's not harmful to change
policy to have external resolver.

Signed-off-by: Sergii Golovatiuk <sgolovatiuk@mirantis.com>

roles/kubernetes/master/templates/manifests/kube-apiserver.manifest.j2

@@ -9,7 +9,7 @@ metadata:
 spec:
   hostNetwork: true
 {% if kube_version | version_compare('v1.6', '>=')  %}
-  dnsPolicy: ClusterFirstWithHostNet
+  dnsPolicy: ClusterFirst
 {% endif %}
   containers:
   - name: kube-apiserver

roles/kubernetes/master/templates/manifests/kube-controller-manager.manifest.j2

@@ -8,7 +8,7 @@ metadata:
 spec:
   hostNetwork: true
 {% if kube_version | version_compare('v1.6', '>=') %}
-  dnsPolicy: ClusterFirstWithHostNet
+  dnsPolicy: ClusterFirst
 {% endif %}
   containers:
   - name: kube-controller-manager

roles/kubernetes/master/templates/manifests/kube-scheduler.manifest.j2

@@ -8,7 +8,7 @@ metadata:
 spec:
   hostNetwork: true
 {% if kube_version | version_compare('v1.6', '>=') %}
-  dnsPolicy: ClusterFirstWithHostNet
+  dnsPolicy: ClusterFirst
 {% endif %}
   containers:
   - name: kube-scheduler

roles/kubernetes/node/templates/manifests/kube-proxy.manifest.j2

@@ -8,7 +8,7 @@ metadata:
 spec:
   hostNetwork: true
 {% if kube_version | version_compare('v1.6', '>=') %}
-  dnsPolicy: ClusterFirstWithHostNet
+  dnsPolicy: ClusterFirst
 {% endif %}
   containers:
   - name: kube-proxy