Docker  

Using Docker and Kubernetes for .NET on Azure AKS and AWS EKS

When your .NET applications outgrow simple PaaS hosting, you’ll likely move toward containerization and orchestration. Containers make your apps portable, and Kubernetes provides the automation for deploying, scaling, and managing them at cloud scale.

In this article, we’ll explore how to:

  • Containerize a .NET app with Docker

  • Deploy it to Azure Kubernetes Service (AKS)

  • Deploy it to Amazon Elastic Kubernetes Service (EKS)

1. Why Containers for .NET?

  • Portability – Runs consistently across dev, test, and cloud.

  • Scalability – Easily scale pods across clusters.

  • Microservices Ready – Break monoliths into smaller services.

  • Cloud-Native – Kubernetes is the industry standard for orchestration.

2. Containerize a .NET App with Docker

Create a new ASP.NET Core app:

dotnet new webapi -o MyKubeApp
cd MyKubeApp

Create a Dockerfile:

# Build stage
FROM mcr.microsoft.com/dotnet/sdk:8.0 AS build
WORKDIR /src
COPY . .
RUN dotnet publish -c Release -o /app

# Runtime stage
FROM mcr.microsoft.com/dotnet/aspnet:8.0
WORKDIR /app
COPY --from=build /app .
ENTRYPOINT ["dotnet", "MyKubeApp.dll"]

Build and run locally:

docker build -t mykubeapp .
docker run -p 8080:80 mykubeapp

3. Deploy to Azure Kubernetes Service (AKS)

Step 1. Create AKS Cluster

az group create --name MyKubeGroup --location eastus
az aks create --resource-group MyKubeGroup --name MyKubeCluster --node-count 2 --enable-addons monitoring --generate-ssh-keys
az aks get-credentials --resource-group MyKubeGroup --name MyKubeCluster

Step 2. Push Image to Azure Container Registry (ACR)

az acr create --resource-group MyKubeGroup --name MyKubeRegistry --sku Basic
az acr login --name MyKubeRegistry
docker tag mykubeapp MyKubeRegistry.azurecr.io/mykubeapp:v1
docker push MyKubeRegistry.azurecr.io/mykubeapp:v1

Step 3. Kubernetes Deployment File (deployment.yaml)

apiVersion: apps/v1
kind: Deployment
metadata:
  name: mykubeapp
spec:
  replicas: 2
  selector:
    matchLabels:
      app: mykubeapp
  template:
    metadata:
      labels:
        app: mykubeapp
    spec:
      containers:
      - name: mykubeapp
        image: mykuberegistry.azurecr.io/mykubeapp:v1
        ports:
        - containerPort: 80
---
apiVersion: v1
kind: Service
metadata:
  name: mykubeapp-service
spec:
  type: LoadBalancer
  selector:
    app: mykubeapp
  ports:
  - protocol: TCP
    port: 80
    targetPort: 80

Apply deployment:

kubectl apply -f deployment.yaml

Your app will be exposed via an Azure Load Balancer IP.

4. Deploy to Amazon Elastic Kubernetes Service (EKS)

Step 1. Create EKS Cluster

eksctl create cluster --name mykubecluster --region us-east-1 --nodes 2

Step 2. Push Image to Amazon Elastic Container Registry (ECR)

aws ecr create-repository --repository-name mykubeapp
$(aws ecr get-login --no-include-email --region us-east-1)
docker tag mykubeapp <account_id>.dkr.ecr.us-east-1.amazonaws.com/mykubeapp:v1
docker push <account_id>.dkr.ecr.us-east-1.amazonaws.com/mykubeapp:v1

Step 3. Apply the Same Kubernetes Manifest

Update the image in deployment.yaml to use your AWS ECR URL, then deploy:

kubectl apply -f deployment.yaml

AWS will provision an ELB (Elastic Load Balancer) and provide a public endpoint.

5. Side-by-Side: AKS vs. EKS

FeatureAKS (Azure)EKS (AWS)
SetupEasy via Azure CLI & PortalRequires eksctl or Terraform, more setup
RegistryAzure Container Registry (ACR)Amazon Elastic Container Registry (ECR)
Load BalancerAzure Load Balancer auto-configuredAWS ELB auto-configured
MonitoringAzure Monitor + Application InsightsCloudWatch + AWS X-Ray
Best ForTeams already using Azure DevOps & App ServicesTeams needing AWS’s global reach & flexibility

6. Best Practices

  • CI/CD Pipelines

    • Azure → GitHub Actions + ACR + AKS

    • AWS → CodePipeline + ECR + EKS

  • Secrets Management

    • Azure Key Vault vs. AWS Secrets Manager

  • Autoscaling: Enable Kubernetes Horizontal Pod Autoscaler (HPA).

  • Observability: Use Prometheus + Grafana, or native cloud monitoring.

  • Multi-Cloud Portability: Keep manifests provider-agnostic, rely on Helm charts or Dapr.

Conclusion

By combining Docker and Kubernetes, your .NET applications become portable, scalable, and ready for production workloads on both Azure AKS and AWS EKS. Azure offers simplicity and deep Microsoft integration, while AWS provides flexibility and global infrastructure.

This dual approach ensures you’re not locked into one cloud and can leverage the best of both worlds.

Membership not found