Skip to content
This repository has been archived by the owner on Jan 12, 2025. It is now read-only.

Latest commit

 

History

History
304 lines (224 loc) · 11 KB

README.md

File metadata and controls

304 lines (224 loc) · 11 KB

Lab 1: Create AKS Cluster

In this lab we will create our Azure Kubernetes Services (AKS) distributed compute cluster.

Prerequisites

  • Azure Account

Instructions

  1. Login to Azure Portal at http://portal.azure.com.

  2. Open the Azure Cloud Shell and choose Bash Shell (do not choose Powershell)

    Azure Cloud Shell

  3. The first time Cloud Shell is started will require you to create a storage account.

  4. Once your cloud shell is started, clone the workshop repo into the cloud shell environment

    git clone https://github.com/Azure/kubernetes-hackfest

    Note: In the cloud shell, you are automatically logged into your Azure subscription.

  5. Ensure you are using the correct Azure subscription you want to deploy AKS to.

    # View subscriptions
az account list

    # Verify selected subscription
az account show

    # Set correct subscription (if needed)
az account set --subscription <subscription_id>

# Verify correct subscription is now set
az account show

  6. Create Azure Service Principal to use through the labs

    az ad sp create-for-rbac --skip-assignment

    This will return the following. !!!IMPORTANT!!! - Please copy this information down as you'll need it for labs going forward.

    "appId": "7248f250-0000-0000-0000-dbdeb8400d85",
"displayName": "azure-cli-2017-10-15-02-20-15",
"name": "http://azure-cli-2017-10-15-02-20-15",
"password": "77851d2c-0000-0000-0000-cb3ebc97975a",
"tenant": "72f988bf-0000-0000-0000-2d7cd011db47"

    Set the values from above as variables (replace and with your values).

    DON'T MESS THIS STEP UP. REPLACE THE VALUES IN BRACKETS!!!

    # Persist for Later Sessions in Case of Timeout
APPID=<appId>
echo export APPID=$APPID >> ~/.bashrc
CLIENTSECRET=<password>
echo export CLIENTSECRET=$CLIENTSECRET >> ~/.bashrc

  7. Create a unique identifier suffix for resources to be created in this lab.

    UNIQUE_SUFFIX=$USER$RANDOM
# Remove Underscores and Dashes (Not Allowed in AKS and ACR Names)
UNIQUE_SUFFIX="${UNIQUE_SUFFIX//_}"
UNIQUE_SUFFIX="${UNIQUE_SUFFIX//-}"
# Check Unique Suffix Value (Should be No Underscores or Dashes)
echo $UNIQUE_SUFFIX
# Persist for Later Sessions in Case of Timeout
echo export UNIQUE_SUFFIX=$UNIQUE_SUFFIX >> ~/.bashrc

    _ Note this value as it will be used in the next couple labs. _

  8. Create an Azure Resource Group in East US.

    # Set Resource Group Name using the unique suffix
RGNAME=aks-rg-$UNIQUE_SUFFIX
# Persist for Later Sessions in Case of Timeout
echo export RGNAME=$RGNAME >> ~/.bashrc
# Set Region (Location)
LOCATION=eastus
# Persist for Later Sessions in Case of Timeout
echo export LOCATION=eastus >> ~/.bashrc
# Create Resource Group
az group create -n $RGNAME -l $LOCATION

  9. Create your AKS cluster in the resource group created above with 3 nodes. We will check for a recent version of kubnernetes before proceeding. We are also including the monitoring add-on for Azure Container Insights. You will use the Service Principal information from step 5.

    Use Unique CLUSTERNAME

    # Set AKS Cluster Name
CLUSTERNAME=aks${UNIQUE_SUFFIX}
# Look at AKS Cluster Name for Future Reference
echo $CLUSTERNAME
# Persist for Later Sessions in Case of Timeout
echo export CLUSTERNAME=aks${UNIQUE_SUFFIX} >> ~/.bashrc

    Get available kubernetes versions for the region. You will likely see more recent versions in your lab.

    az aks get-versions -l $LOCATION --output table

KubernetesVersion    Upgrades
-------------------  ------------------------
1.14.0(preview)      None available
1.13.5               1.14.0(preview)
1.12.8               1.13.5
1.12.7               1.12.8, 1.13.5
1.11.9               1.12.7, 1.12.8
1.11.8               1.11.9, 1.12.7, 1.12.8
1.10.13              1.11.8, 1.11.9
1.10.12              1.10.13, 1.11.8, 1.11.9
1.9.11               1.10.12, 1.10.13
1.9.10               1.9.11, 1.10.12, 1.10.13

    Set the version to one with available upgrades (in this case v 1.12.8)

    K8SVERSION=1.13.5

    The below command can take 10-20 minutes to run as it is creating the AKS cluster. Please be PATIENT and grab a coffee...

    # Create AKS Cluster
az aks create -n $CLUSTERNAME -g $RGNAME \
--kubernetes-version $K8SVERSION \
--service-principal $APPID \
--client-secret $CLIENTSECRET \
--generate-ssh-keys -l $LOCATION \
--node-count 3 \
--enable-addons monitoring \
--no-wait

  10. Verify your cluster status. The ProvisioningState should be Succeeded

    az aks list -o table
    Name             Location    ResourceGroup            KubernetesVersion    ProvisioningState    Fqdn
---------------  ----------  -----------------------  -------------------  -------------------  ----------------------------------------------------------------
aksstephen14260  eastus      aks-rg-stephen14260      1.12.8               Succeeded            aksstephen-aks-rg-stephen14-62afe9-9aa48ae4.hcp.eastus.azmk8s.io

  11. Get the Kubernetes config files for your new AKS cluster

    az aks get-credentials -n $CLUSTERNAME -g $RGNAME

  12. Verify you have API access to your new AKS cluster

    Note: It can take 5 minutes for your nodes to appear and be in READY state. You can run watch kubectl get nodes to monitor status.

    kubectl get nodes
    NAME                       STATUS   ROLES   AGE     VERSION
aks-nodepool1-14089323-0   Ready    agent   113s    v1.12.8
aks-nodepool1-14089323-1   Ready    agent   2m59s   v1.12.8
aks-nodepool1-14089323-2   Ready    agent   2m1s    v1.12.8

    To see more details about your cluster:

    kubectl cluster-info
    Kubernetes master is running at https://aksstephen-aks-rg-stephen14-62afe9-9aa48ae4.hcp.eastus.azmk8s.io:443
Heapster is running at https://aksstephen-aks-rg-stephen14-62afe9-9aa48ae4.hcp.eastus.azmk8s.io:443/api/v1/namespaces/kube-system/services/heapster/proxy
CoreDNS is running at https://aksstephen-aks-rg-stephen14-62afe9-9aa48ae4.hcp.eastus.azmk8s.io:443/api/v1/namespaces/kube-system/services/kube-dns:dns/proxy
kubernetes-dashboard is running at https://aksstephen-aks-rg-stephen14-62afe9-9aa48ae4.hcp.eastus.azmk8s.io:443/api/v1/namespaces/kube-system/services/kubernetes-dashboard/proxy
Metrics-server is running at https://aksstephen-aks-rg-stephen14-62afe9-9aa48ae4.hcp.eastus.azmk8s.io:443/api/v1/namespaces/kube-system/services/https:metrics-server:/proxy

    You should now have a Kubernetes cluster running with 3 nodes. You do not see the master servers for the cluster because these are managed by Microsoft. The Control Plane services which manage the Kubernetes cluster such as scheduling, API access, configuration data store and object controllers are all provided as services to the nodes.

Namespaces Setup

This lab creates namespaces that reflect a representative example of an organization's environments. In this case dev, uat and prod. We will also apply the appopriate permissions, limits and resource quotas to each of the namespaces.

  1. Navigate to the directory of the cloned repository

    cd kubernetes-hackfest

  2. Create three namespaces

    # Create namespaces
kubectl apply -f labs/create-aks-cluster/create-namespaces.yaml

# Look at namespaces
kubectl get ns

  3. Assign CPU, memory and storage limits to namespaces

    # Create namespace limits
kubectl apply -f labs/create-aks-cluster/namespace-limitranges.yaml

# Get list of namespaces and drill into one
kubectl get ns
kubectl describe

  4. Assign CPU, Memory and Storage Quotas to Namespaces

    # Create namespace quotas
kubectl apply -f labs/create-aks-cluster/namespace-quotas.yaml

# Get list of namespaces and drill into one
kubectl get ns
kubectl describe ns dev

  5. Test out Limits and Quotas in dev Namespace

    # Test Limits - Forbidden due to assignment of CPU too low
kubectl run nginx-limittest --image=nginx --restart=Never --replicas=1 --port=80 --requests='cpu=100m,memory=256Mi' -n dev

# Test Limits - Pass due to automatic assignment within limits via defaults
kubectl run nginx-limittest --image=nginx --restart=Never --replicas=1 --port=80 -n dev

# Check running pod and dev Namespace Allocations
kubectl get po -n dev
kubectl describe ns dev

# Test Quotas - Forbidden due to memory quota exceeded
kubectl run nginx-quotatest --image=nginx --restart=Never --replicas=1 --port=80 --requests='cpu=500m,memory=1Gi' -n dev

# Test Quotas - Pass due to memory within quota
kubectl run nginx-quotatest --image=nginx --restart=Never --replicas=1 --port=80 --requests='cpu=500m,memory=512Mi' -n dev

# Check running pod and dev Namespace Allocations
kubectl get po -n
kubectl describe n

  6. Clean up limits, quotas, pods

    kubectl delete -f labs/create-aks-cluster/namespace-limitranges.yaml
kubectl delete -f labs/create-aks-cluster/namespace-quotas.yaml
kubectl delete po nginx-limittest nginx-quotatest -n dev

kubectl describe ns dev
kubectl describe ns uat
kubectl describe ns prod

  7. Create namespace for our application. This will be used in subsequent labs.

    kubectl create ns hackfest

Troubleshooting / Debugging

  • The limits and quotas of a namespace can be found via the kubectl describe ns <...> command. You will also be able to see current allocations.
  • If pods are not deploying then check to make sure that CPU, Memory and Storage amounts are within the limits and do not exceed the overall quota of the namespace.

Docs / References

Next Lab: Build Application Components and Prerequisites