Vanilla K8s in Vagrant

latest K8s install (v1.28 at the time) using containerd runtime on Vagrant nodes

Vagrant part

run vagrant up in root directory to provision a single master Node (control-plane)
- by default 3 (three) worker Nodes will be provisioned (can be modified with WorkerNodeCount variable)
- master-provision.sh is used to configure the control-plane node and worker-provision.sh for worker Nodes
- master-node has needed tools + some additional testing/debugging tools and binaries for administrative work

connect to master-node using vagrant ssh master-node
run sudo -i and inside /root/k8s/ directory run ./create-cluster-admin.sh
- this will configure Admin user for cluster and install all needed components
  - Flannel CNI
  - Metrics Server
  - Istio ingress + istioctl
  - MetalLB
[OPTIONAL]: if you'd like to manage the cluster from host using your tools run ./converge-kube-config.sh to update local ./kube/config file
- k9s and kubectl are also installed in master-node so cluster can be configured from there directly
join worker Nodes using the token provided with first script

use manifests in k8s/manifests directory per your liking, everything provided as learning examples

[ISSUE]: Vagrant network sometimes is known to hang with "protocol not supported" error and prevents guest-to-guest communication and even host-to-guest one
- only ICMP works at that moment
- can prevent joining worker Nodes to k8s cluster
- can prevent managing the cluster from host
- [FIX]: run sudo ip neigh flush all on master-node to clear ARP cache and try again