This lab provides an introduction to Docker along with Kubernetes. It also shows how to create a Kubernetes (K8s) Cluster using ACS along with how to deploy different types of workloads.
The purpose of this section is to help you get familar with Kubernetes on Azure. There are a couple of options, Azure Container Service (ACS) which allows for complete control of the environment, and our new Managed Kubernetes Service (AKS) which is in PREVIEW. We are going to focus on AKS as that is the future and allows us to focus on deploying Applications versus managing Infrastructure.
- To make this easier, we are going to leverage the existing Tutorial in Azure Docs. Click here to navigate to the Docs.
- The first step is to create the Cluster.
- The second step is to connect to the Cluster.
In this section we will leverage https://KataCoda.com and use their interactive exercises.
- Click here and do the following exercises:
Deploy Containers Using Kubectl
The purpose of this section is similar to the Web App for Containers first challenge, show how easy it is to get started with Containers in Azure. The difference is that this is ACI which feels more like Serverless, and bills like Serverless, versus Web App for Containers which is PaaS.
- Spin up a Container using ACI via the az cli that is exposed via a Public IP. The az command you want to look at is container. A sample command should start with
az containerand then add the necessary parameters. One thing to keep in mind, pick a Web Application Image as we are gonig to expose it on a port so it needs to be HTTP accessible.
Hint: Click here if you are having difficulties.
The purpose of this section is to help you understand how to deploy Containers to Managed Kubernetes Services (AKS), which you will realize is exactly the same as Azure Container Service (ACS) Kubernetes.
- Create Sample Application
- Initially the EXTERNAL-IP for the azure-vote-front service appears as pending. Once the EXTERNAL-IP address has changed from pending to an IP address put that in your browser to test out the app.
# Run K8s File.
kubectl create -f azure-vote.yaml
# Test the App.
kubectl get service azure-vote-front --watch- Cleanup Sample Application
- Remove the Deployment from Kubernetes Cluster
# Delete the external facing front-end service.
kubectl delete svc azure-vote-front
# Run the Test line again to see what has changed.
# Delete the external facing back-end service.
kubectl delete svc azure-vote-back
# Delete the deployment.
kubectl delete deploy azure-vote-frontThe purpose of this section is to help you understand a couple of the key building blocks when it comes to deploying Containers in Kubernetes. There are sample files in this directoy to help you out with the following challenges.
- Create a Pod manifest file that has the following parameters pod.yaml:
- Uses 2 Labels zone = prod and version = v1
- Uses the evillgenius\kuar:1 image
- Exposes port 8080
- Verify Pod is operational by using
kubectl port-forwardcommand to forward port 8080 on your local host to port 8080 on the pod and checking with you browserhttp://localhost:8080
- Create a Deployment manifest using the same parameters as above but add deploy.yaml:
- You will need to replace
<APP_NAME_GOES_HERE> - Make 3 replicas of the app
- Use a RollingUpdate strategy with no less than 1 pod avaialble and no more than 1 pod extra during updates
- Verify deployment is operational by using
kubectl port-forwardcommand to forward port 8080 on your local host to port 8080 on any of the pods in the deployment and checking with you browserhttp://localhost:8080
- You will need to replace
- Create a Service manifest that exposes the Deployment above using a LoadBalancer service.yaml:
- Hint: You need to link the Service and Deployment together.
- Note: Getting an external IP may take some time initially. use
kubectl get service -wuntil an external IP is shown. - Test Container via Public IP
The purpose of this section is to help you combine all of the concepts we have covered throughout the day by deploying your application from your own Private Regsitry, in this case ACR.
- Summary Steps Througout the Day:
- Pull in code from a git repository
- Add/Copy the code into your container via dockerfile
- Build a new Container Image
- Push the new image to a Private Container Registry (ACR)
- Deploy the new container via K8s
- Combine the Deployment Manifest and the Service Manifest yaml files from previous step into a single Kubernetes Manifest yaml file (example]and deploy it into a new Namespace using Container Images from your Private Registry.
- Hint: Take a look at the azure-vote.yaml file that was used to deploy the full application as an example of how to combine the different K8s building blocks together and pull from ACR.
Hint: If you are struggling, you can leverage the existing Tutorial in Azure Docs. Click here to navigate to the Docs.
- Some versions of Azure-CLI will not pull down/copy your
./kube/configfile properly to your local machine- This is a known bug and has a fix already, but may not be available yet depeding on your OS's system packge for Azure-CLI or your installation method for Azure-CLI
- Solution: you must pull the
.kube/configfile down manually
# go to your ~/.kube folder
cd ~/.kube
# copy the remote config file to your local ~/.kube folder
scp -i /path/to/your/ssh/key <your_kube_user>@<kube_cluster_master_name>.<azure_location>.cloudapp.azure.com:~/.kube/config ~/.kube
# example
scp -i ~/.ssh/id_rsa azureuser@rkk8sclust-resourcegroup-uuid.canadacentral.cloudapp.azure.com.canadacentral.cloudapp.azure.com:~/.kube/config ~/.kube
Once the config file is in your ~/.kube directory you will be able to run the kubectl command against your remote cluster.
Note: The .kube/config file defines the cluster host information including the SSH keys to use to connect to the remote cluster/master.
putty.exe -ssh -i <path to private key file> -L 80:localhost:80 <User name>@<Public DNS name of your master node>