This test will fail if the playbook is not running on the master node, because it tries to access url on `localhost`. Installation doesn't verify any of the other prometheus components are running, so I believe this verification can be removed.
@@ -48,15 +48,6 @@
-n "{{ openshift_prometheus_namespace }}"
| {{ openshift_client_binary }} apply --config={{ mktemp.stdout }}/admin.kubeconfig -f - -n "{{ openshift_prometheus_namespace }}"
-- name: Verify that node exporter is running
- uri:
- url: http://localhost:9100/metrics
- register: node_exporter_health
- until: node_exporter_health.status == 200
- retries: 120
- delay: 1
- changed_when: false
-
- name: Remove temp directory
file:
state: absent