Prometheus + Kubernetes demo

Step 1: Deploy

Helm is the easiest way to do this. Check out https://github.com/kubernetes/charts/tree/master/stable/prometheus.

We can run helm install stable/prometheus to get the stock prometheus server.

In this case, we will run: helm install stable/prometheus --name prom-demo -f values.yaml to use our custom yaml.

Step 2: See it running

Running minikube service prom-demo-prometheus-server will bring up the browser with prometheus server running.

Step 3: Check out Kubernetes stats

Check out kube_pod_status_phase to see our all our pods running.

Step 4: Prometheus operators

More info here: https://prometheus.io/docs/prometheus/latest/querying/operators/ but you can use the query language to clean up the data point we looked at above.

sum(kube_pod_status_phase) by (phase) will return the set of pods that are running grouped by their phases.

Step 5: Check out node information

The node exporter gives you node relative information as well like CPU/disk usage etc.

Run machine_cpu_cores will return the CPU count which should match the number of CPU's in kubectl describe nodes

Step 6: App metrics

1. Run the app: kubectl apply -f twelve-clouds.yaml
2. Visit the app after it's deployed: minikube service twelve-clouds-service
3. You'll see the /metrics endpoint with go stats
4. Visit the /hello endpoint 2 times.
5. You'll see a new "hello_calls" metric in the dashboard