With our new namespace available, it's time to create a very simple Prometheus configuration and save it on a ConfigMap using the following manifest:

apiVersion: v1
kind: ConfigMap
metadata:
  name: prometheus-config
  namespace: monitoring 
data:
  prometheus.yml: |
    scrape_configs:
    - job_name: prometheus
      static_configs:
      - targets:
        - localhost:9090

Apply the previous manifest using the following command:

kubectl apply -f prometheus-configmap.yaml

Now, it's time to start a new deployment of Prometheus, making sure we mount the previously configured ConfigMap into the pod we are deploying. The Deployment object is configured with the following metadata:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: prometheus-deployment
  namespace: monitoring
  labels:
    app: prometheus

The Prometheus container will be started with its configuration file and data directory coming from volume mounts, shown as follows:

        args:
          - --config.file=/etc/config/prometheus.yml
          - --storage.tsdb.path=/data
        volumeMounts:
          - name: config-volume
            mountPath: /etc/config/prometheus.yml
            subPath: prometheus.yml
          - name: prometheus-data
            mountPath: /data
            subPath: ""

The config-volume volume is created from a ConfigMap, while the prometheus-data volume is created with an empty directory. This can be seen in the following snippet:

      volumes:
        - name: config-volume
          configMap:
           name: prometheus-config
        - name: prometheus-data
          emptyDir: {}

Apply the previous manifest using the following command:

kubectl apply -f prometheus-deployment.yaml

We can follow the deployment status using this snippet:

kubectl rollout status deployment/prometheus-deployment -n monitoring

We should look at the logs of our Prometheus instance using the following command:

kubectl logs --tail=20 -n monitoring -l app=prometheus

After a successful deployment, we're ready to assign a new service to our instance, choosing NodePort so we can access it without requiring port-forwarding, like so:

kind: Service
apiVersion: v1
metadata:
  name: prometheus-service
  namespace: monitoring
spec:
  selector:
    app: prometheus
  type: NodePort
  ports:
  - name: prometheus-port
    protocol: TCP
    port: 9090
    targetPort: 9090

Apply the previous manifest using the following:

kubectl apply -f prometheus-service.yaml

And you're ready to check your new Prometheus service using the following code snippet:

minikube service prometheus-service -n monitoring

This will open your browser on the Prometheus service endpoint. You can now check the running configuration and targets using the Prometheus web interface:

Now that we have Prometheus running in Kubernetes, we can start adding targets for it to scrape. In the next section, we will have a look at how you can achieve this.