Increase the delay between checking for image pull success

Also, ignore the failure. We're really just trying to prime things so
that when pods start the start up faster. If the image pull is a failure
then we'll similarly fail to start the pods.

roles/openshift_control_plane/tasks/main.yml

@@ -152,6 +152,8 @@
   until: job_result.finished
   when: control_plane_image.stdout_lines == []
   retries: 30
+  delay: 10
+  failed_when: false
 
 - name: Check status of etcd image pre-pull
   async_status:
@@ -163,6 +165,8 @@
   - "'stdout_lines' in etcd_image_exists"
   - etcd_image_exists.stdout_lines == []
   retries: 30
+  delay: 10
+  failed_when: false
 
 - name: Start and enable self-hosting node
   systemd:

roles/openshift_node/tasks/config.yml

@@ -60,3 +60,5 @@
   - node_image.stdout_lines == []
   - not openshift_is_atomic | bool
   retries: 30
+  delay: 10
+  failed_when: false

roles/openshift_node/tasks/node_system_container.yml

@@ -25,6 +25,8 @@
   - node_image is defined
   - node_image.stdout_lines == []
   retries: 30
+  delay: 10
+  failed_when: false
 
 - name: Copy node container image to ostree storage
   command: >