Skip to content

doc.md

Latest commit

 

History

History
182 lines (122 loc) · 5.57 KB

doc.md

File metadata and controls

182 lines (122 loc) · 5.57 KB

kind load docker-image bluerise:latest --name my-cluster

How I deployed my nodejs weather App into Kubernetes

This setup replicates a deployment of an enterprise microservice. I used (Kubernetes IN Docker) kubernetes cluster, easy for low payload launches into kubernetes. Therefore ensure you have docker desktop on your engine when you want to use KIND The application to be lauched here is a custom application previously built as a microservices app.

Deploying an Application into A kubernetes cluster in easy steps.

  1. Install KIND Click Kind to install.

  2. Create a Kubernetes CLUSTER Run

 kind create cluster --name bluerise

alt text

Interact with the new cluster a bit

kubectl cluster-info --context kind-bluerise
  1. LOAD image of the application/s for your workflow
kind load docker-image johnstx/bluerise:v1.3
  1. DEPLOY the application manifest files. (Deployments, Services etc)

bluerise-deployment.yaml -

apiVersion: apps/v1
kind: Deployment
metadata:
  name: bluerise-deployment
spec:
  replicas: 1
  selector:
    matchLabels:
      app: bluerise
  template:
    metadata:
      labels:
        app: bluerise
    spec:
      containers:
      - name: bluerise-weather-app
        image: johnstx/bluerise:v1
        ports:
        - containerPort: 3000


kubectl apply -f bluerise-deployment.yaml

Manifest file for the service

bluerise-service.yaml

apiVersion: v1
kind: Service
metadata:
  name: bluerise-service
spec:
  type: NodePort
  selector:
    app: bluerise
  ports:
    - protocol: TCP
      port: 80
      targetPort: 3000
      nodePort: 30081

kubectl apply -f bluerise-service.yaml

Check out the resources for more info using kubectl get

alt text

alt text

Access the application through the browser

Port-forward the service to enable http access.

kubectl port-forward svc/bluerise 8080:80

alt text

Accessing the app on then browser

alt text

helm repo add prometheus-community https://prometheus-community.github.io/helm-charts

helm install monitor prometheus-community/kube-prometheus-stack -n monitoring

NAME: monitors LAST DEPLOYED: Sun Jun 15 00:18:50 2025 NAMESPACE: monitoring STATUS: deployed REVISION: 1 NOTES: kube-prometheus-stack has been installed. Check its status by running: kubectl --namespace monitoring get pods -l "release=monitors"

Get Grafana 'admin' user password by running:

kubectl --namespace monitoring get secrets monitors-grafana -o jsonpath="{.data.admin-password}" | base64 -d ; echo

Access Grafana local instance:

export POD_NAME=$(kubectl --namespace monitoring get pod -l "app.kubernetes.io/name=grafana,app.kubernetes.io/instance=monitors" -oname) kubectl --namespace monitoring port-forward $POD_NAME 3000

Visit https://github.com/prometheus-operator/kube-prometheus for instructions on how to create & configure Alertmanager and Prometheus instances using the Operator.

BUGS After installing the helm release, only the node exporter pod was running, the other thre pods were stuck on "containerCreating for a while", then the metrics container came on leaving the operator and grafana pods -- All came on but one pod for the monitors-grafana.

Create a ServiceMonitor to explicitly tell Prometheus to scrape the service.

Step 3: Access Prometheus and Grafana Dashboards

📊 Port-forward Prometheus: kubectl port-forward svc/monitors-kube-prometheus-s-prometheus -n monitoring 9090 Open: http://localhost:9090/targets

Confirm that bluerise appears in the "targets" list and is marked UP.

kubectl port-forward svc/monitors-grafana -n monitoring 3000:80

Prometheus failed to show any metrics from the app.?

This is because he endpoint in the nodejs application config was not specified. So I have to revisit the bluerise application, and the pipeline for docker build. This is where you'd appreciate CI/CD. So we are going to modify the index.js file to include the prom-client and /metrics endpoint for prometheus to scrap info from it. Then using jenkins to build the image.

  1. install prom-client
  2. modify the index.js file to specify the client, prometheus setup and the metrics endpoint for prometheus.

New Bug -

/metrics endopoint EMPTY!!

Bug fix

Check that package.json fine lists prom-client as a dependency. NB The node_modules directory should also list prom-client by now, after the install.

Test the app and check the /metrics endpoint in your browser. You should find a records on this page.

success. OK

Then push to your repo and pipeline.

http://localhost:9090/targets run this to see the targets prometheus scrapes

Bug

Servicemonitor for my app was not listed above, but could be seen when k get servicemonitors

Bugfix

describe on app's servicemonitor revealed the it was not exactly scraping any source, as noticed from the port section of the code which should have the name of the port as http (it is supposeed to scraping the app service) of the app's service directory. HOwever, the service doesnt have a name specified as http, so it leaves an empty targt for the servicemonitor. Solution is to modify the port field of the service by adding name:http to it.