Deploy .NET Web Application In Kubernetes

Introduction

In this article, I will explain the basics of how to deploy a .NET Web application in Kubernetes. I will use Minikube to deploy the app on the local machine.

You can either use an existing .NET Web Application or create a new .NET Web Application and containerize it.

I am going to use the app that was created in my previous article.

I’ve made some minor adjustments for clarity and proper capitalization. Containerize Your .NET 7.0 Web Application With Docker

Kubernetes (k8s) can only deploy containers because it only knows how to deploy containers.

In the previous article, I have already containerized the application using Docker.

There are many other tools to containerize the application, such as LXC, Podman, Containerd, etc. However, Docker is the most popular tool for creating and running containers.

Prerequisites

You need to install following tools in your machine:

1. Docker
2. Kubectl
3. Minikube

Create Docker Image

Fist, step to deploy app in Kubernetes is build the image of web app that you want to run in Kubernetes.

I am building image form Containerize Your .NET 7.0 Web Application With Docker

Go inside project folder.

1

cd dockerizeDotNet7App/dockerizeDotNet7App 

Run command in terminal to build image:

1

docker build -t dockerizedotnet7app-web_app .

Sharing Docker image with a registry

There are many registry to publish docker images. Here, I am using Docker Hub to publish image publicly.

Login to Docker Hub

Use below command to login to the Docker Hub. This command will help you to login in the Docker Hub.

1

docker login

Push the Image to the Registry:

Push the image to the registry:

1

docker push <your-registry-username>/<your-repository-name:tag>

In my case, my docker push command will be like:

1

docker push thebhandariprakash/dockerizedotnet7app-web_app

Deploying Your App in Kubernetes

I will deploy an application in Mac OS minikube local development Kubernetes cluster.

1. Deployment (deployment.yaml)

A deployment allows you to describe an application’s life cycle, such as which images to use for the app, the number of pods that we want. Update pods and replica sets, rollback to previous deployment versions, scale a deployment , pause or continue a deployment, etc

We define set of instructions in a YAML file. ie deployment.yaml

Now, you can apply the deployment.yaml file by using below command:

1

kubectl apply -f deployment.yaml

Once, deployment is created you will get deployment.apps/web-app created success message.

To verify, you can run the below command, which will list the pods.

1

 kubectl get pods

The number of pods is depend upon the replicas value in deployment.yaml file replicas: 1. In you case replicas value is 1.

1
2

NAME                      READY   STATUS              RESTARTS   AGE
web-app-586b96bbb-2rjkf   1/1     Running             0          10s

Even if you delete pods with below command. It will spin up another pod immediately.

1
2
3
4
5

kubectl delete pod <pod name>

Example : 
kubectl delete pod web-app-586b96bbb-2rjkf
pod "web-app-586b96bbb-2rjkf" deleted

Again, if you run kubectl get pods, you will see new pod.

2. Service (service.yaml)

Services provide a way to expose your application’s Pods to the network or other services within the cluster, and they play a very important role in enabling communication between different parts of your application.

Here is a service.yaml file that I am using:

Now, you can create the service by using service.yaml. If we run below command it will crate service:

1

kubectl create -f service.yaml

Once, service is created you will see service/web-app created success message.

To verify, you can run the below command, which will list the pods.

1

 kubectl get services

Now, you can see created service in your terminal.

1
2
3

NAME         TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)        AGE
kubernetes   ClusterIP   10.96.0.1       <none>        443/TCP        41h
web-app      NodePort    10.96.157.175   <none>        80:31747/TCP   4s

3. Accessing an app via browser

To check the app via browser or form any HTTP client. First you need to find the URL in which app is accessible.

In minikube you can run following command.

minikube service --url=true dotnet7-web-app

The above command will give the base URL. Something like:

1
2

http://127.0.0.1:636660
❗  Because you are using a Docker driver on darwin, the terminal needs to be open to run it.

Now, you are ready to access you app via browser. For my app I can use http://127.0.0.1:63660/WeatherForecast to see the response.

4. Create an Ingress and use custom domain to access App via browser

Ingress is important in k8s it exposes HTTP and HTTPS routes from outside the cluster to services within the cluster. Traffic routing is controlled by rules defined on the Ingress resource.

The Ingress is the equivalent of a reverse proxy in Kubernetes.

Everytime running this command minikube service --url=true dotnet7-web-app and getting the app running URL is tedious.

So, I have created the ingress.yml file.

Here is a ingress.yaml file that I am going to use:

In this file I have added, host as kubernete.local ie. - host: kubernete.local. Which is the custom domain. App should be accessible in this custom domain http://kubernete.local

By default, minikube comes with the Nginx as Ingress controller, and you can enable it with below command:

minikube addons enable ingress

To enable ingress addons Minikube will take few minutes because it takes time to download and install Nginx as an Ingress.

Once ingress addons is ready, we can create Ingress with below command:

1

kubectl create -f ingress.yaml

Now, you can find the Minikube IP with below command:

minikube ip

Let’s say you have your minikube IP address 192.168.49.2

Now, you need add my host names to /etc/hosts

1

192.168.49.2   kubernete.local

If you try http://kubernete.local/WeatherForecast in your browser, it should connect to app and give the result back.

Here is my output form custom domain http://kubernete.local/WeatherForecast

If you are using M1 with the docker driver. You need to enable minikube tunnel. Keep in mind that your etc/hosts file needs to map to 127.0.0.1, instead of the output of minikube ip or kubectl get ingress This is an very important.

Conclusion

This article has provided a guide to deploying a .NET Web application in Kubernetes. Throughout the process, I’ve covered essential steps from creating Docker images to setting up Kubernetes deployments and services. By utilizing tools like Minikube, Docker, and Kubernetes, I’ve successfully containerized and deployed the application for local Minikube Kubernetes. Additionally, I have explored advanced concepts such as creating an Ingress to enable custom domain access.