If you want to manually create the cluster (i.e., without GKE) using kubeadm, then this document may help you to do it. Before moving into the procedure it is good to have knowledge on kubeadm.
Prerequisites:
OS: Ubuntu version 16.04 or higher
Minimum Memory required: For master node - 2 CPU's, 8GB memory
For each worker node - 1 CPU's, 4GB memory
Let us look into the steps for cluster creation using kubeadm
Step 1: Docker Installation
Create the Instances in the GCE , It should be at least one master node and two or more worker nodes. Follow these steps for all the nodes.
Docker installation is the primary steps in the cluster creation, so follow the below steps for all the nodes you created. Otherwise you can also look into the official docker documentation for docker installation.
$ sudo apt-get update -y
$ sudo apt-get install -qy docker.io
Check whether docker got properly installed or not using docker --version command
Step 2: Install Kubeadm
Get Kubernetes repo key
$ curl -s https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo apt-key add
Add Kubernetes repository to manifests
$ sudo apt-add-repository "deb http://apt.kubernetes.io/ kubernetes-xenial main"
Install Kubeadm
$ sudo apt install kubeadm
Check kubeadm version and make sure it is properly installed or not.
$ kubeadm version
Step 3: Create the cluster
It is possible to configure kubeadm init with a configuration file instead of command line flags, and some more advanced features may only be available as configuration file options. This file is passed in the --config option.
On master node:
Initialize the kubeadm
$ sudo kubeadm init [option]
Options:
--pod-network-cidr=10.244.0.0/16
--config=/root/kubeadm-config.yaml
--cert-dir
Copy and execute kubectl config file
$ mkdir -p $HOME/.kube
$ sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
$ sudo chown $(id -u):$(id -g) $HOME/.kube/config
Step 4: Join worker node
To join worker nodes to the master node, run the below command separately on each worker node
As a root user
$ kubeadm join 10.156.0.18:6443 --token ow15j0.kz12nltctqeowkiy \
> --discovery-token-ca-cert-hash sha256:a7024eacb754a01721f28cedc52e92427a83225db0f800d1bfb9117f2832602c
To check whether all the pods are active and running, run the following command on the master node.
$ kubectl get nodes
Step 5: Troubleshooting
When you are initializing kubeadm with cidr, then If you get pods status as not ready then run the following code in master node
$ sudo kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml
If you are doing kubeadm init with config file then sample template of the file is shown below.
Kubeadm-config.yaml sample template
You can access the yaml file here.
Tear down:
To tear down the kubeadm, run these commands on the master node
$ kubectl drain <node name> — delete-local-data — force — ignore-daemonsets
$ kubectl delete node <node name>
Then, on the node being removed, reset all kubeadm installed state:
$ kubeadm reset
