This repository setup the Kubernetes cluster with kubeadm via ansible.

This repository have some restrictions now. you can see below Restrictions.

• join control plane nodes.

Now the repository only support one control plane node and multi workers, it will support multi control plane nodes late.

• os: CentOS

• CRI: docker

• CNI: flannel

CentOS 7.4+

Install the Kubernetes cluster via ansible, so you should install ansible in your machine.

Ansible installation guide you can open ansible offical document.

This repository now only support CentOS 7.4+, so you install ansible with yum.

/bin/bash install.sh

Add your hosts into inventory/hosts.ini, format like:

k8s-1 ansible_ssh_host=1.2.3.4 ansible_ssh_port=22 ansible_ssh_user=root ansible_ssh_pass=123456

You can choose which hosts to use as the master role, and which hosts to use as the node role.

kube-masters install components, include: kube-apiserver, kube-scheduler, kube-controller-manager, etcd

kube-nodes install components, include: kubelet.

all hosts install docker, kube-proxy, kube-flannel.

Ansible variables in inventory/group_vars/all.yml.

ansible-playbook -i inventory/hosts.ini playbooks/deploy.yml

When you run setup command successfully, you can use kubectl get nodes to show your cluster all nodes, and run kubectl get pod -n kube-system to show all pods in you cluster.

This is a Kubernetes cluster that have one control plane(master) and two workers.

The inventory/hosts.ini is like below:

k8s-1 ansible_ssh_host=192.168.46.11 ansible_ssh_port=22 ansible_ssh_user=root ansible_ssh_pass=123456
k8s-2 ansible_ssh_host=192.168.46.12 ansible_ssh_port=22 ansible_ssh_user=root ansible_ssh_pass=123456
k8s-3 ansible_ssh_host=192.168.46.13 ansible_ssh_port=22 ansible_ssh_user=root ansible_ssh_pass=123456

[kube-masters]
k8s-1

[kube-nodes]
k8s-2
k8s-3

[k8s-cluster:children]
kube-masters
kube-nodes

[etcd:children]
kube-masters

If cluster created successfully, you can see these nodes and below pods in kube-system namespace.

[root@k8s-1 ~]# kubectl get nodes
NAME    STATUS   ROLES    AGE     VERSION
k8s-1   Ready    master   15h     v1.14.1
k8s-2   Ready    <none>   15h     v1.14.1
k8s-3   Ready    <none>   15h     v1.14.1

[root@k8s-1 ~]# kubectl get pod -n kube-system
NAME                            READY   STATUS    RESTARTS   AGE
coredns-fb8b8dccf-6sqlr         1/1     Running   0          15h
coredns-fb8b8dccf-jr6n7         1/1     Running   0          15h
etcd-k8s-1                      1/1     Running   0          15h
kube-apiserver-k8s-1            1/1     Running   0          15h
kube-controller-manager-k8s-1   1/1     Running   0          15h
kube-flannel-ds-amd64-5l5r8     1/1     Running   0          15h
kube-flannel-ds-amd64-fz7f5     1/1     Running   0          15h
kube-flannel-ds-amd64-ll4g5     1/1     Running   0          15h
kube-proxy-m2tx8                1/1     Running   0          15h
kube-proxy-tssfg                1/1     Running   0          15h
kube-proxy-zcf4z                1/1     Running   0          15h
kube-scheduler-k8s-1            1/1     Running   0          15h