* Upgrade kubernetes to v1.13.0 * Remove all precense of scheduler.alpha.kubernetes.io/critical-pod in templates * Fix cert dir * Use kubespray v2.8 as baseline for gitlab
* Switch to kubeadm deployment mode Discuss:https://github.com/kubernetes-incubator/kubespray/issues/3301 * Add non-kubeadm upgrage to kubeadm cluster
* Upgrade kubernetes to v1.12.0 Use kubeadm v1alpha3 config * Upgrade coredns and etcd * Upgrage docker to 18.06