Skip to content

Latest commit

 

History

History
136 lines (108 loc) · 3.55 KB

azure-kubeadm.md

File metadata and controls

136 lines (108 loc) · 3.55 KB
Raw

Install k8s cluster on azure VMs using kubeadm

*local.* files contains sensitive data so they have not been pushed on the remote repo. The equivalent file contains the value placeholder to be replaced w/ your personal values.

  • provision azure resources from local shell
bash ./azure-configure-master.sh

from k8s-master1 shell

sudo cp ~/cloud.conf /etc/kubernetes
  • create the cluster using kubeadm and configure master node
bash k8sMaster.sh | tee ~/master.out

OR

  • create the cluster using kubeadm

from k8s-master1 shell

sudo kubeadm init --config kubeadm.conf

output should be the following

Your Kubernetes master has initialized successfully!

To start using your cluster, you need to run the following as a regular user:

mkdir -p $HOME/.kube sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config sudo chown $(id -u):$(id -g) $HOME/.kube/config

You should now deploy a pod network to the cluster. Run "kubectl apply -f [podnetwork].yaml" with one of the options listed at: https://kubernetes.io/docs/concepts/cluster-administration/addons/

You can now join any number of machines by running the following on each node as root:

kubeadm join XXXX:6443 --token XXXX --discovery-token-ca-cert-hash > sha256:XXXX

  • configure kubectl config from k8s-master1 shell
mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config
  • install Canal CNI pluging from k8s-master1 shell
wget https://docs.projectcalico.org/v3.3/getting-started/kubernetes/installation/hosted/canal/rbac.yaml -O rbac.yaml
wget https://docs.projectcalico.org/v3.3/getting-started/kubernetes/installation/hosted/canal/canal.yaml  -O canal.yaml

kubectl apply -f rbac.yaml
kubectl apply -f canal.yaml

  • join the worker node(s) from worker node(s) shell use the kubeadm join... saved in master.out

  • download k8s configuration

scp "cristiano@master-1-$RESOURCE_GROUP.westeurope.cloudapp.azure.com:~/.kube/config" k8s.local.conf

then run the latest bash interactive not the standard login

exec bash

  • modify the k8s.local.conf replacing internal mater node ip with public dns as configured in kubeadm.conf

  • deploy a basic ngnix app from local shell

kubectl --kubeconfig k8s.local.conf create -f basic.yaml

OR

from k8s-master1 shell

kubectl create -f basic.yaml

OR a more complex app

kubectl apply -f https://raw.githubusercontent.com/Azure-Samples/azure-voting-app-redis/master/azure-vote-all-in-one-redis.yaml
  • get pod and service from k8s-master1 shell
kubectl get pod,svc
  • make sure canal has been installed on all nodes
kubectl get pods -n kube-system -o wide | grep canal
  • test ClusterIP on all nodes from k8s nodes shell
curl http://<basicservice ClusterIP>
  • test NodePort on all nodes from local shell
NODE_PORT=<NodePort>
curl http://master-1-$RESOURCE_GROUP.westeurope.cloudapp.azure.com:$NODE_PORT
curl http://worker-1-$RESOURCE_GROUP.westeurope.cloudapp.azure.com:$NODE_PORT
curl http://worker-2-$RESOURCE_GROUP.westeurope.cloudapp.azure.com:$NODE_PORT
  • if you'd like to deploy application workload on master node too, you have to remove the taint that prevents it
kubectl taint nodes --all node-role.kubernetes.io/master:NoSchedule-