Get Your First Application on Kubernetes
A quick and easy tutorial for getting a Node.js application on Kubernetes.
Join the DZone community and get the full member experience.Join For Free
This article talks about the technical details on deploying your first application on Kubernetes.
What Is This Application?
This is a very simple Node.js application. This application exposes a GET endpoint "/", and when invoked will return "Hello World" message!
Though this example is simple, this will help in understanding all the basic stuff that is needed to run a Dockerized application on Kubernetes.
The following are the pre-requisites to run this application. Refer to their official site to install these, if you do not have them already.
- Node: https://nodejs.org/en/
- Docker: https://docs.docker.com/install/
- Dockerhub account.
- Kubernetes: https://kubernetes.io/docs/setup/pick-right-solution/#local-machine-solutions: Use the packages like "minikube" or "microk8s for Ubuntu" for easier setup.
Steps to Follow
cd nodejs-docker-hw docker build -t abhishekjv/nodejs-hw .
docker build command will refer to the Dockerfile in present working directory, and create the Docker image. This Docker image is the portable executable application that can run on a machine that has Docker and even on the production environments like Kubernetes, Amazon ECS, OpenShift etc.
-t option for the
docker build command is used to give a name to the Docker image. If this option is not specified, a random name will be provided which might not be friendly! When tagging the image with the name, it is recommended to use the format "<REPOSITORY-ID>/<IMAGE-NAME>."
- REPOSITORY ID is the Docker hub ID or the private registry ID you may be using.
- IMAGE-NAME is the simple name of the image.
Here, the image is tagged as "abhishekjv/nodejs-hw."
2.1: Quick look at Dockerfile
FROM node:4.6 WORKDIR /app ADD . /app RUN npm install EXPOSE 3000 CMD npm start
FROM node:4.6: Says what is the source image needed to run this application.
WORKDIR /app: Describes what should be the working directory of the application inside the image.
ADD . /app : Tells to put the contents of the current directory into the working directory. Current directory is the location where Dockerfile is located.
npm install : Tells Docker to run this command once files are copied. This will install the npm dependencies.
EXPOSE 3000: Tells Docker to expose port 3000 from the container. (This port is also defined in index.js)
npm start : Finally, Docker will execute the command defined here to start the application.
Verify if the image is built by executing Docker images command:
docker images $ docker images
Execute the below command to run the Docker image locally.
docker run -p 3000:3000 929654e7274a
The host system port 3000 is being mapped to the container port 3000.
Go to http://localhost:3000 to verify the working app
docker push abhishekjv/nodejs-hw
kubectl create -f pod-nodejs-docker-hw.yml
This will create a pod in K8s setup. Verify the status by executing the command kubectl get pod
6. To Access the Application Deployed On K8s:
Once the application is deployed to K8s cluster, the port of K8s has to be exposed to host machine to access the application. This can be done in 2 ways in local environment: (When running in AWS or other production setup, it is recommended to configure and use load balancers)
kubectl port-forward nodejs-docker-hw 8081:3000 Forwarding from 127.0.0.1:8081 3000 Forwarding from [::1]:8081 300
The port 3000 of the container is exposed on port 8081 on the host system.
Go to http://localhost:8081 to access the application.
kubectl expose pod nodejs-docker-hw --type=NodePort --name nodejs-hw-service
To get the port number:
kubectl describe service nodejs-hw-service
Look for the port number defined for NodePort:
NodePort: <unset> 30469/TCP
Use the above port from local system to access the application: http://localhost:30469
Finally, here is the link to Git repo.
Published at DZone with permission of Abhishek Jawali. See the original article here.
Opinions expressed by DZone contributors are their own.