Description of problem:
When running the logging checks via openshift-ansible it's possible to encounter following traceback.


Version-Release number of the following components:
I ran the installation with the latest pre-release ose-ansible build v3.6.137-1

rpm -q openshift-ansible: openshift-ansible-3.6.137-1.git.0.12654fb.el7.noarch
rpm -q ansible: ansible-2.2.3.0-1.el7.noarch
FWIW I installed against CentOS 7, containerized with the Origin 3.6.0-alpha.1 images specified.


How reproducible:
100%

Steps to Reproduce:
1. Deploy cluster with logging.
2. Break a component so that its pod(s) are defined but not scheduled. One way that probably works is to mark all nodes unschedulable and delete an existing pod.
3. Run the logging checks.

Expected results:
Should report on broken components

Actual results:
Will attach privately. Traceback is as follows:
Traceback (most recent call last):
  File "/usr/lib/python2.7/site-packages/ansible/executor/task_executor.py", line 125, in run
    res = self._execute()
  File "/usr/lib/python2.7/site-packages/ansible/executor/task_executor.py", line 522, in _execute
    result = self._handler.run(task_vars=variables)
  File "/home/lmeyer/go/src/github.com/openshift/openshift-ansible/roles/openshift_health_checker/action_plugins/openshift_health_check.py", line 66, in run
    r = check.run(tmp, task_vars)
  File "/home/lmeyer/go/src/github.com/openshift/openshift-ansible/roles/openshift_health_checker/openshift_checks/logging/kibana.py", line 39, in run
    check_error = self.check_kibana(kibana_pods)
  File "/home/lmeyer/go/src/github.com/openshift/openshift-ansible/roles/openshift_health_checker/openshift_checks/logging/kibana.py", line 105, in check_kibana
    not_running = self.not_running_pods(pods)
  File "/home/lmeyer/go/src/github.com/openshift/openshift-ansible/roles/openshift_health_checker/openshift_checks/logging/logging.py", line 59, in not_running_pods
    for container in pod['status']['containerStatuses']
KeyError: 'containerStatuses'