- Development Environment Setup
- Containerize the application
- Deploy the application to local Kubernetes Cluster
- Push the image to Azure Container Registry using the Azure CLI
- Create AKS cluster using the Azure CLI
- Deploy the application to AKS Cluster using .yml file
- Scale the application and Kubernetes infrastructure
- Update the application then push the image to ACR and update deployment
Development Environment Setup
- Install Docker
- Enable Kubernetes Cluster on Docker, under Docker -> Settings
- Install Azure CLI and Azure PowerShell modules
Containerize the application
- Clone this repo to a folder
- Open a command prompt and navigate to your project folder.
- Use the following commands to build and run your Docker image:
docker build . -t kubeservice:local
docker run -d -p 8000:80 kubeservice:local
- View the .Net Core application running from a container by navigating to localhost:8000
Deploy the application to local Kubernetes Cluster
-
Deploy docker image to Kubernetes
kubectl run kubeservice-deployment --image=kubeservice:local --port=80 --replicas=3
-
kubectlcommand- to verify the deployment:
kubectl get deployments - to verify the replicasets:
kubectl get rs - to verify the Pods:
kubectl get pods
- to verify the deployment:
- Expose the application to outside world, use the following command
kubectl expose deployment kubeservice-deployment --type=NodePort
- Verify the service
kubectl get svc
Expected Result:
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
kubernetes ClusterIP 10.0.0.1 <none> 443/TCP 42d
kubeservice-deployment NodePort 10.0.120.27 <none> 80:31057/TCP 1m
-
Finally access the app using browser: http://localhost:31057
Push the image to Azure Container Registry using the Azure CLI
- Login to Azure
az login
- Create Azure Resource Group
az group create -n vvkubeserviceacr-rg -l eastus
- Create Azure Container Registry
az acr create -n vvkubeserviceacr -g vvkubeserviceacr-rg -l eastus --sku standard
-
- Verify the Azure Container Registry
- Verify the Azure Container Registry
az acr list -o table
Expected Result:
NAME RESOURCE GROUP LOCATION SKU LOGIN SERVER CREATION DATE ADMIN ENABLED
vvkubeserviceacr vvkubeserviceacr-rg eastus Standard vvkubeserviceacr.azurecr.io 2018-10-27T20:04:22Z
- Push the image to Azure Container Registry
- Login to Azure Container Registry
az acr login -n vvkubeserviceacr
-
- Tag the docker image prior to push the image to Azure Container Registry
docker tag kubeservice:local vvkubeserviceacr.azurecr.io/kubeservice:v1
-
- Push the tagged image to Azure Container Registry
docker push vvkubeserviceacr.azurecr.io/kubeservice:v1
-
- Verify the repository in Azure Container Registry
az acr repository list -n vvkubeserviceacr -o table
Result:
kubeservice
Create AKS cluster using the Azure CLI
AKS uses the Service Principal to access other Azure Services like the Azure Container Registry (ACR). It is recommended to register an application in Azure AD and create an identity for the application (i.e. Service Principal)
-
Create Service Principal
az ad sp create-for-rbac --skip-assignment
Result:
{
"appId": "<appId>",
"displayName": "<displayName>",
"name": "<name>",
"password": "<password>",
"tenant": "<tenant>"
}
- Create acrID variable
$acrID = az acr show --name vvkubeserviceacr --resource-group vvkubeserviceacr-rg --query "id" --output tsv
- Create a role assignment with appid and acrID
az role assignment create --assignee "<appId>" --role Reader --scope $acrID
- Create AKS Cluster
az aks create --name vvkubeserviceaks --resource-group vvkubeserviceacr-rg --node-count 1 --generate-ssh-keys --service-principal "<appId>" --client-secret "<password>" --location eastus
- Get the AKS cluster credentials to update local environment
az aks get-credentials --name vvkubeserviceaks --resource-group vvkubeserviceacr-rg
- Verify the AKS Cluster in Azure Cloud
kubectl get nodes
Expected Result look like this:
NAME STATUS ROLES AGE VERSION
aks-nodepool1-33154602-0 Ready agent 6m v1.9.11
Deploy the application to AKS Cluster using .yml file
- Below is the content of
kubeservice-deploy.ymlfile:
apiVersion: apps/v1
kind: Deployment
metadata:
name: kubeservice-deployment
labels:
app: kubeservice
spec:
replicas: 3
template:
metadata:
name: kubeservice
labels:
app: kubeservice
spec:
containers:
- name: kubeservice
image: vvkubeserviceacr.azurecr.io/kubeservice:v1
imagePullPolicy: IfNotPresent
restartPolicy: Always
selector:
matchLabels:
app: kubeservice
---
apiVersion: v1
kind: Service
metadata:
name: kubeservice-service
spec:
selector:
app: kubeservice
ports:
- port: 80
type: LoadBalancer
- Deploy the application to AKS Cluster
kubectl apply -f .\kubeservice-deploy.yml
- To monitor the service deployment
kubectl get service --watch
Result:
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
kubernetes ClusterIP 10.0.0.1 <none> 443/TCP 20m
kubeservice-service LoadBalancer 10.0.154.43 <pending> 80:30845/TCP 15s
kubeservice-service LoadBalancer 10.0.154.43 137.135.124.83 80:30845/TCP 1m
Scale the application and Kubernetes infrastructure
Scaling Azure Kubernetes Service (AKS) Cluster nodes
- Get the current nodes
kubectl get nodes
Result:
NAME STATUS ROLES AGE VERSION
aks-nodepool1-33154602-0 Ready agent 26m v1.9.11
- Scaling nodes
az aks scale --resource-group vvkubeserviceacr-rg --name=vvkubeserviceaks --node-count 3
Result:
<json with success message>
- Get the current nodes after scaling
kubectl get nodes
Result:
NAME READY STATUS RESTARTS AGE
kubeservice-deployment-5fbf677667-ffftl 1/1 Running 0 17m
kubeservice-deployment-5fbf677667-kvfg2 1/1 Running 0 17m
kubeservice-deployment-5fbf677667-mjgbq 1/1 Running 0 17m
Scaling Azure Kubernetes Service (AKS) Cluster pods
- Verify the current replicas # of application
kubectl get pods
Result:
NAME READY STATUS RESTARTS AGE
kubeservice-deployment-5fbf677667-ffftl 1/1 Running 0 18m
kubeservice-deployment-5fbf677667-kvfg2 1/1 Running 0 18m
kubeservice-deployment-5fbf677667-mjgbq 1/1 Running 0 18m
- Verify the current deployment
kubectl get deployment
Result:
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
kubeservice-deployment 3 3 3 3 19m
- Scale the pods to 5
kubectl scale --replicas=5 deployment/kubeservice-deployment
Result:
deployment.extensions "kubeservice-deployment" scaled
- Verify the current deployment after scale
kubectl get deployment
Result:
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
kubeservice-deployment 5 5 5 5 22m
- Verify the current replicas # of application
kubectl get pods
Result:
NAME READY STATUS RESTARTS AGE
kubeservice-deployment-5fbf677667-b6p8l 1/1 Running 0 2m
kubeservice-deployment-5fbf677667-ffftl 1/1 Running 0 22m
kubeservice-deployment-5fbf677667-kvfg2 1/1 Running 0 22m
kubeservice-deployment-5fbf677667-mjgbq 1/1 Running 0 22m
kubeservice-deployment-5fbf677667-qdrsf 1/1 Running 0 2m
Update the application then push the image to ACR and update deployment
-
Change the application
Change the label "Host Name:" with "AKS Host Name:"
-
Create a new image
docker build . -t vvkubeserviceacr.azurecr.io/kubeservice:v2 -
Push the image to Azure Container Registry
docker push vvkubeserviceacr.azurecr.io/kubeservice:v2
- Update the deployment
kubectl set image deployment kubeservice-deployment kubeservice=vvkubeserviceacr.azurecr.io/kubeservice:v2
- To monitor the deployment
kubectl get pods
Result:
NAME READY STATUS RESTARTS AGE
kubeservice-deployment-9c594f98-5bwgp 1/1 Running 0 1m
kubeservice-deployment-9c594f98-b4szp 1/1 Running 0 1m
kubeservice-deployment-9c594f98-ctbzp 1/1 Running 0 1m
kubeservice-deployment-9c594f98-pjztm 1/1 Running 0 1m
kubeservice-deployment-9c594f98-w75g4 1/1 Running 0 1m
- Test the updated application
kubectl get service
Result:
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
kubernetes ClusterIP 10.0.0.1 <none> 443/TCP 1h
kubeservice-service LoadBalancer 10.0.154.43 137.135.124.83 80:30845/TCP 1h
-
Run the application to see the changes made by navigating to the url: http://137.135.124.83
