Operators were introduced by CoreOS as a class of software that operates other software, putting operational knowledge collected by humans into software. Read more in the original blog post.
The mission of the Prometheus Operator is to make running Prometheus on top of Kubernetes as easy as possible, while preserving configurability as well as making the configuration Kubernetes native.
To follow this getting started you will need a Kubernetes cluster you have access to. Let's give the Prometheus Operator a spin:
apiVersion: rbac.authorization.k8s.io/v1alpha1 kind: ClusterRoleBinding metadata: name: prometheus-operator roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: prometheus-operator subjects: - kind: ServiceAccount name: prometheus-operator namespace: default --- apiVersion: rbac.authorization.k8s.io/v1alpha1 kind: ClusterRole metadata: name: prometheus-operator rules: - apiGroups: - extensions resources: - thirdpartyresources verbs: - create - apiGroups: - monitoring.coreos.com resources: - alertmanagers - prometheuses - servicemonitors verbs: - "*" - apiGroups: - apps resources: - statefulsets verbs: ["*"] - apiGroups: [""] resources: - configmaps - secrets verbs: ["*"] - apiGroups: [""] resources: - pods verbs: ["list", "delete"] - apiGroups: [""] resources: - services - endpoints verbs: ["get", "create", "update"] - apiGroups: [""] resources: - nodes verbs: ["list", "watch"] --- apiVersion: v1 kind: ServiceAccount metadata: name: prometheus-operator --- apiVersion: extensions/v1beta1 kind: Deployment metadata: name: prometheus-operator labels: operator: prometheus spec: replicas: 1 template: metadata: labels: operator: prometheus spec: serviceAccountName: prometheus-operator containers: - name: prometheus-operator image: quay.io/coreos/prometheus-operator:v0.8.1 resources: requests: cpu: 100m memory: 50Mi limits: cpu: 200m memory: 100Mi
The Prometheus Operator introduces third party resources in Kubernetes to declare the desired state of a Prometheus and Alertmanager cluster as well as the Prometheus configuration. The resources it introduces are:
Important for this guide are the
ServiceMonitorresources. Have a look at the alerting guide for more information about the
Alertmanagerresource or the design doc for an overview of all resources introduced by the Prometheus Operator.
Prometheus resource declaratively describes the desired state of a Prometheus deployment, while a
ServiceMonitor describes the set of targets to be monitored by Prometheus.
Prometheus resource includes a selection of
ServiceMonitors to be used, this field is called the
First, deploy three instances of a simple example application, which listens and exposes metrics on port
apiVersion: extensions/v1beta1 kind: Deployment metadata: name: example-app spec: replicas: 3 template: metadata: labels: app: example-app spec: containers: - name: example-app image: fabxc/instrumented_app ports: - name: web containerPort: 8080
ServiceMonitor has a label selector to select
Services and the underlying
Endpoints objects. The
Service object for the example application selects the
Pods by the
app label having the
example-app value. In addition to that the
Service object specifies the port the metrics are exposed on.
kind: Service apiVersion: v1 metadata: name: example-app labels: app: example-app spec: selector: app: example-app ports: - name: web port: 8080
Service object is discovered by a
ServiceMonitor, which selects in the same way. The
app label must have the value
apiVersion: monitoring.coreos.com/v1alpha1 kind: ServiceMonitor metadata: name: example-app labels: team: frontend spec: selector: matchLabels: app: example-app endpoints: - port: web
Prometheus object defines the
serviceMonitorSelector to specify which
ServiceMonitors should be included. Above the label
team: frontend was specified, so that's what the
Prometheus object selects by.
apiVersion: monitoring.coreos.com/v1alpha1 kind: Prometheus metadata: name: example spec: serviceMonitorSelector: matchLabels: team: frontend version: v1.6.1 resources: requests: memory: 400Mi
This way the frontend team can create new
Services resulting in
Prometheus to be dynamically reconfigured.
To be able to access the Prometheus instance it will have to be exposed to the outside somehow. For demonstration purpose it will be exposed via a
Service of type
apiVersion: v1 kind: Service metadata: name: prometheus-example spec: type: NodePort ports: - name: web nodePort: 30900 port: 9090 protocol: TCP targetPort: web selector: prometheus: example
Service is created the Prometheus web UI is available under the node's IP address on port
30900. The targets page in the web UI now shows that the instances of the example application have successfully been discovered.
Exposing the Prometheus web UI may not be an applicable solution. Read more about the possibilities of exposing it in the exposing Prometheus and Alertmanager guide.