A playbooks to building the hard way Kubernetes cluster, This playbook is a fully automated command to bring up a Kubernetes cluster on VM or Baremetal.

Feature list:

• Support Kubernetes v1.10.0+.
• Highly available Kubernetes cluster.
• Full of the binaries installation.
• Kubernetes addons:
  • Promethues monitoring.
  • Metrics Server.
  • EFK logging.
  • Ingress Controller.
  • Kubernetes Dashboard.
• Support container network:
  • calico.
  • flannel.
• Support container runtime:
  • docker.
  • nvidia-docker.(Require NVIDIA driver and CUDA 9.0+)
  • containerd.
  • cri-o.

In this section you will deploy a cluster using vagrant.

Prerequisites:

• Ansible version: v2.5 (or newer).
• Vagrant: >= 2.0.0.
• VirtualBox: >= 5.0.0.
• Mac OS X need to install sshpass tool.

$ brew install http://git.io/sshpass.rb

The getting started guide will use Vagrant with VirtualBox to deploy a Kubernetes cluster on virtual machines. You can deploy the cluster with a single command:

$ ./hack/setup-vms
Cluster Size: 1 master, 2 worker.
  VM Size: 1 vCPU, 2048 MB
  VM Info: ubuntu16, virtualbox
  CNI binding iface: eth1
Start to deploy?(y):

• You also can use sudo ./hack/setup-vms -p libvirt -i eth1 command to deploy the cluster on KVM.

If you want to access API you need to create RBAC object define the permission of role. For example using cluster-admin role:

$ kubectl create clusterrolebinding open-api --clusterrole=cluster-admin --user=system:anonymous

Login the addon's dashboard:

• Dashboard: https://API_SERVER:8443/api/v1/namespaces/kube-system/services/https:kubernetes-dashboard:/proxy/
• Logging: https://API_SERVER:8443/api/v1/namespaces/kube-system/services/kibana-logging/proxy/

As of release 1.7 Dashboard no longer has full admin privileges granted by default, so you need to create a token to access the resources:

$ kubectl -n kube-system create sa dashboard
$ kubectl create clusterrolebinding dashboard --clusterrole cluster-admin --serviceaccount=kube-system:dashboard
$ kubectl -n kube-system get sa dashboard -o yaml
apiVersion: v1
kind: ServiceAccount
metadata:
  creationTimestamp: 2017-11-27T17:06:41Z
  name: dashboard
  namespace: kube-system
  resourceVersion: "69076"
  selfLink: /api/v1/namespaces/kube-system/serviceaccounts/dashboard
  uid: 56b880bf-d395-11e7-9528-448a5ba4bd34
secrets:
- name: dashboard-token-vg52j

$ kubectl -n kube-system describe secrets dashboard-token-vg52j
...
token:      eyJhbGciOiJSUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJrdWJlcm5ldGVzL3NlcnZpY2VhY2NvdW50Iiwia3ViZXJuZXRlcy5pby9zZXJ2aWNlYWNjb3VudC9uYW1lc3BhY2UiOiJrdWJlLXN5c3RlbSIsImt1YmVybmV0ZXMuaW8vc2VydmljZWFjY291bnQvc2VjcmV0Lm5hbWUiOiJkYXNoYm9hcmQtdG9rZW4tdmc1MmoiLCJrdWJlcm5ldGVzLmlvL3NlcnZpY2VhY2NvdW50L3NlcnZpY2UtYWNjb3VudC5uYW1lIjoiZGFzaGJvYXJkIiwia3ViZXJuZXRlcy5pby9zZXJ2aWNlYWNjb3VudC9zZXJ2aWNlLWFjY291bnQudWlkIjoiNTZiODgwYmYtZDM5NS0xMWU3LTk1MjgtNDQ4YTViYTRiZDM0Iiwic3ViIjoic3lzdGVtOnNlcnZpY2VhY2NvdW50Omt1YmUtc3lzdGVtOmRhc2hib2FyZCJ9.bVRECfNS4NDmWAFWxGbAi1n9SfQ-TMNafPtF70pbp9Kun9RbC3BNR5NjTEuKjwt8nqZ6k3r09UKJ4dpo2lHtr2RTNAfEsoEGtoMlW8X9lg70ccPB0M1KJiz3c7-gpDUaQRIMNwz42db7Q1dN7HLieD6I4lFsHgk9NPUIVKqJ0p6PNTp99pBwvpvnKX72NIiIvgRwC2cnFr3R6WdUEsuVfuWGdF-jXyc6lS7_kOiXp2yh6Ym_YYIr3SsjYK7XUIPHrBqWjF-KXO_AL3J8J_UebtWSGomYvuXXbbAUefbOK4qopqQ6FzRXQs00KrKa8sfqrKMm_x71Kyqq6RbFECsHPA

Copy and paste the token to dashboard.

In this section you will manually deploy a cluster on your machines.

Prerequisites:

• Ansible version: v2.5 (or newer).
• Linux distributions: Ubuntu 16+/Debian/CentOS 7.x.
• All Master/Node should have password-less access from deploy node.

For machine example:

Add the machine info gathered above into a file called inventory/hosts.ini. For inventory example:

[etcds]
k8s-m1
k8s-n[1:2]

[masters]
k8s-m1
k8s-n1

[nodes]
k8s-n[1:3]

[kube-cluster:children]
masters
nodes

Set the variables in group_vars/all.yml to reflect you need options. For example:

# overide kubernetes version(default: 1.10.6)
kube_version: 1.11.2

# container runtime, supported: docker, nvidia-docker, containerd.
container_runtime: docker

# container network, supported: calico, flannel.
cni_enable: true
container_network: calico
cni_iface: ''

# highly available variables
vip_interface: ''
vip_address: 172.16.35.9

# etcd variables
etcd_iface: ''

# kubernetes extra addons variables
enable_dashboard: true
enable_logging: false
enable_monitoring: false
enable_ingress: false
enable_metric_server: true

# monitoring grafana user/password
monitoring_grafana_user: "admin"
monitoring_grafana_password: "p@ssw0rd"

If everything is ready, just run cluster.yml playbook to deploy the cluster:

$ ansible-playbook -i inventory/hosts.ini cluster.yml

And then run addons.yml to create addons:

$ ansible-playbook -i inventory/hosts.ini addons.yml

Verify that you have deployed the cluster, check the cluster as following commands:

$ kubectl -n kube-system get po,svc

NAME                                 READY     STATUS    RESTARTS   AGE       IP             NODE
po/haproxy-master1                   1/1       Running   0          2h        172.16.35.10   k8s-m1
...

Finally, if you want to clean the cluster and redeploy, you can reset the cluster by reset-cluster.yml playbook.:

$ ansible-playbook -i inventory/hosts.ini reset-cluster.yml

Pull requests are always welcome!!! I am always thrilled to receive pull requests.