Asp.Net core + Kubernetes

This project is a self learning for first timers. The objective of this project is to create a simple asp.net core web application using the standard dotnet template and then delpoy it to a kubernetes cluster. The plan is for learning and see it working on local system.

Pre-Requisite

- Windows 10
- Code Editor(Visual Studio Code)
- Docker Desktop for Windows
- MiniKube
- dotnet core

Installation

Before starting we have to make sure your environment is ready. Let's start by installing the VS Code.

Tables
install VSCode

Install the latest dotnet sdk

Tables
install dotnet

Install Docker Desktop

Tables
install docker

Install Minikube

It's easier to install minikube with all its dependencies using chocaltey. Once installed run the below command in a command prompt, preferably in Admin mode.

choco install minikube

Creating a dotnet application.

For this example we have used the dotnet angular template. Open VSCode in your working folder and create the sample dotnet project.

dotnet new angular -o learning

Check your application by running it from vs code. Hopefully it should work.

Preparing your application for docker.

Before you create a docker file to deploy your application, we have to change our application a bit. Open the learning.csproj file and remove the PublishSPAWebpack Target section

<Target Name="PublishRunWebpack" AfterTargets="ComputeFilesToPublish"> <!-- As part of publishing, ensure the JS resources are freshly built in production mode --> <Exec WorkingDirectory="$(SpaRoot)" Command="npm install" /> <Exec WorkingDirectory="$(SpaRoot)" Command="npm run build -- --prod" /> <Exec WorkingDirectory="$(SpaRoot)" Command="npm run build:ssr -- --prod" Condition=" '$(BuildServerSideRenderer)' == 'true' " /> <!-- Include the newly-built files in the publish output --> <ItemGroup> <DistFiles Include="$(SpaRoot)dist\**; $(SpaRoot)dist-server\**" /> <DistFiles Include="$(SpaRoot)node_modules\**" Condition="'$(BuildServerSideRenderer)' == 'true'" /> <ResolvedFileToPublish Include="@(DistFiles->'%(FullPath)')" Exclude="@(ResolvedFileToPublish)"> <RelativePath>%(DistFiles.Identity)</RelativePath> <CopyToPublishDirectory>PreserveNewest</CopyToPublishDirectory> <ExcludeFromSingleFile>true</ExcludeFromSingleFile> </ResolvedFileToPublish> </ItemGroup> </Target>

We will understand later why we need to remove this step from csproj.

Create Docker file

To deploy our application we need to prepare our application for docker. We need to create the docker file.

You can grab the sample docker file from docker, however since we have used dotnet core 3.1 we have to do some minor changes. Check my docker file in my git to find the changes.

Make sure to change the second param of the ENTRYPOINT array to the name of your DLL.

Build your Image

docker build -t learning .

You can see why we had to change the docker file for build. We use two environments to build our application as we use dotnet core environment to build the dotnet and use a node environment to build the angular.

Your docker image is built successfully and ready for deployment. You can choose your environment and at this point deploy to a docker and see if it's working.

Deploy your Image

Make sure your minikube is started.

minikube start --container-runtime docker --vm-driver=docker

To deploy your image you need a yaml file. You can read more about yaml here

You can find my yaml at my git

All you say in your yaml is your container name and few more metadata to it. kubectl should have been installed with minikube.

kubectl create -f deployment.yaml

Now check if the deployment has succeeded

kubectl get deployments

You can see that though your deployment has gone through they are not yet available. It's because by default minikube does not know about our image available on docker daemon. We have to share the context for it to know the availability of the image or alternatively push your image to docker hub.

 & minikube -p docker3 docker-env | Invoke-Expression

If you check we have put imagePullPolicy: Never. This is to ensure that the Docker daemon doesn’ t try to pull unknown images from the Docker registry.

If you re-run your command

kubectl get deployments 

You should be able to see that the deployments is working. All good so now we know that the image is created and deployed successfully. You can fire the minikube dashboard to see it.

minikube dashboard

Now how to do we see the same in our browser. For that to work we have to expose the application as a service. The step is similar we have to deploy a service exposing our application. You can check my [services.yaml])(https://github.com/PankajPatro/learning-docker-kubernetes/blob/master/services.yaml)

kubectl create -f deployment.yaml

Once you see that your services is deployed successfully. All you have to do is find the url and open the url in your browser to see your application in action.

Get your service url

minikube service learning --url

Only required to be done in windows as minikube has to tunnel

Run your service

minikube service learning

Once you see your web application. You can do your clean up

kubectl delete service learning
kubectl delete deployment learning
minikube delete
minikube stop

Enjoy coding. Please visit my git