Bug 1744752: Add a new alert rule for reporting the Machine-api Operator failure

install/0000_90_machine-api-operator_04_alertrules.yaml

@@ -4,11 +4,11 @@ metadata:
   labels:
     prometheus: k8s
     role: alert-rules
-  name: machineapioperator-rules
+  name: machine-api-operator-prometheus-rules
   namespace: openshift-machine-api
 spec:
   groups:
-    - name: general.rules
+    - name: machine-without-valid-node-ref
       rules:
         - alert: MachineWithoutValidNode
           expr: |
@@ -18,25 +18,23 @@ spec:
             severity: critical
           annotations:
             message: "machine {{ $labels.name }} does not have valid node reference"
-
----
-apiVersion: monitoring.coreos.com/v1
-kind: PrometheusRule
-metadata:
-  labels:
-    prometheus: k8s
-    role: alert-rules
-  name: machineapioperator-rules
-  namespace: openshift-machine-api
-spec:
-  groups:
-    - name: general.rules
+    - name: machine-api-operator-metrics-collector-up
       rules:
-        - alert: MachineMAOMetricsDown
+        - alert: MachineAPIOperatorMetricsCollectionFailing
           expr: |
              mapi_mao_collector_up == 0
           for: 3m
           labels:
             severity: critical
           annotations:
-            message: "machine api MAO metric {{ $labels.kind }} is not being reported"
+            message: "machine api operator metrics collection is failing. For more details:  oc logs <machine-api-operator-pod-name> -n openshift-machine-api"
+    - name: machine-api-operator-down
+      rules:
+        - alert: MachineAPIOperatorDown
+          expr: |
+             absent(up{job="machine-api-operator"} == 1)
+          for: 3m
+          labels:
+            severity: critical
+          annotations:
+            message: "machine api operator is down"