Monitoring for ironic-prometheus-exporter

This commit adds the changes to enable monitoring
in the machine-api-operator that will allow Prometheus
to collect data from the ironic-prometheus-exporter[1] that
runs in the ironic-image [2].

- Added monitoring label to namespace yaml
- Added monitoring information to the rbac yaml
- Added Service for the ironic-prometheus-exporter
- Added the ServiceMonitor for the ironic-prometheus-exporter
- Added PrometheusRule with alerts for baremetal_temp_celsius
metric.

[1] https://github.com/metal3-io/ironic-prometheus-exporter
[2] https://github.com/metal3-io/ironic-image

install/0000_30_machine-api-operator_00_namespace.yaml

@@ -7,3 +7,4 @@ metadata:
   labels:
     name: openshift-machine-api
     openshift.io/run-level: "1"
+    openshift.io/cluster-monitoring: "true"

install/0000_30_machine-api-operator_09_rbac.yaml

@@ -269,3 +269,21 @@ subjects:
   - kind: ServiceAccount
     name: machine-api-operator
     namespace: openshift-machine-api
+
+---
+- apiVersion: rbac.authorization.k8s.io/v1
+  kind: Role
+  metadata:
+    name: prometheus-k8s
+    namespace: openshift-monitoring
+  rules:
+  - apiGroups:
+    - ""
+    resources:
+    - services
+    - endpoints
+    - pods
+    verbs:
+    - get
+    - list
+    - watch

install/0000_30_machine-api-operator_13_ironic-exporter-service.yaml

@@ -0,0 +1,21 @@
+---
+apiVersion: monitoring.coreos.com/v1
+kind: Service
+metadata:
+  name: metal3-baremetalhost-controller
+  namespace: openshift-machine-api
+  labels:
+    app: ironic-exporter
+spec:
+  ports:
+    - name: http
+      protocol: TCP
+      port: 9608
+      targetPort: 9608
+  selector:
+    app: ironic-exporter
+  clusterIP: None
+  type: ClusterIP
+  sessionAffinity: None
+status:
+  loadBalancer: {}

install/0000_30_machine-api-operator_14_ironic-exporter-servicemonitor.yaml

@@ -0,0 +1,21 @@
+---
+apiVersion: monitoring.coreos.com/v1
+kind: ServiceMonitor
+metadata:
+  labels:
+    app: ironic-exporter
+  name: metal3-baremetalhost-controller
+  namespace: openshift-machine-api
+spec:
+  endpoints:
+  - port: "9608-tcp"
+    scheme: http
+    path: /metrics
+    targetPort: 9608
+  jobLabel: app
+  namespaceSelector:
+    matchNames:
+    - metal3-baremetalhost-controller
+  selector:
+    matchLabels:
+      app: ironic-exporter

install/0000_30_machine-api-operator_15_ironic-exporter-prometheusrules.yaml

@@ -0,0 +1,25 @@
+apiVersion: monitoring.coreos.com/v1
+kind: PrometheusRule
+metadata:
+  name: metal3-baremetalhost-controller
+  namespace: openshift-machine-api
+spec:
+  groups:
+  - name: metal3-baremetalhost-controller
+    rules:
+    - alert: HighCPUTemperature
+      annotations:
+        summary: "The baremetal node {{ $labels.node_name }} CPU {{ $labels.entity_id }} is too high"
+        description: "The baremetal node {{ $labels.node_name }} CPU {{ $labels.entity_id }} is too high in the past minute. Last measurement {{ $value }}"
+      expr: baremetal_temp_celsius > 96
+      for: 5m
+      labels:
+        severity: warning
+    - alert: LowCPUTemperature
+      annotations:
+        summary: "The baremetal node {{ $labels.node_name }} CPU {{ $labels.entity_id }} is too low"
+        description: "The baremetal node {{ $labels.node_name }} CPU {{ $labels.entity_id }} is too low in the past minute. Last measurement {{ $value }}"
+      expr: baremetal_temp_celsius < 3
+      for: 5m
+      labels:
+        severity: warning