added script and readme

README.md

@@ -0,0 +1,29 @@
+# Kubernetes Cluster Deployment Guide
+
+This guide describes how to deploy a Kubernetes cluster tailored to your environment using the provided `init-deploy.sh` script.
+
+## ⚙️ Pre-requisites
+
+Before running the deployment script, ensure the following are in place:
+
+- `kubectl` is installed on your local machine
+- The remote user can execute `sudo` commands **without a password prompt**
+- SSH host keys of the target nodes are already added to your system (to avoid interactive confirmation prompts)
+
+## 🧭 Deployment Goals
+
+This setup will create a Kubernetes cluster with the following topology and components:
+
+- **1 Control Plane Node**
+- **2 Worker Nodes**
+- **Calico** as the CNI (Container Network Interface)
+- **MetalLB** as the cloud load balancer for bare metal services
+- **Longhorn** as the persistent storage provider
+
+## 🚀 How to Deploy
+
+To initiate the deployment, run the provided script:
+
+```bash
+chmod +x init-deploy.sh
+./init-deploy.sh

ansible/k8s-init.yaml

@@ -2,6 +2,7 @@
   hosts: k8s_control_plane
   become: true
   vars:
+    ## This is the default networkd of Calico
     pod_network_cidr: "192.168.0.0/16"
     apiserver_advertise_address: "10.50.3.21"
     cri_socket: "unix:///run/containerd/containerd.sock"

ansible/longhorn-reqs.yaml

@@ -0,0 +1,36 @@
+---
+- name: Setup Longhorn Requirements
+  hosts: all
+  become: true
+  tasks:
+
+    - name: Install required packages
+      apt:
+        name:
+          - open-iscsi
+          - nfs-common
+        state: present
+        update_cache: yes
+
+    - name: Ensure iscsi_tcp kernel module is loaded
+      modprobe:
+        name: iscsi_tcp
+        state: present
+
+    - name: Ensure iscsi_tcp module loads on boot
+      copy:
+        dest: /etc/modules-load.d/iscsi.conf
+        content: |
+          iscsi_tcp
+        owner: root
+        group: root
+        mode: '0644'
+
+    - name: Disable and stop multipathd service
+      systemd:
+        name: multipathd
+        enabled: false
+        state: stopped
+        masked: true
+        daemon_reload: yes
+      ignore_errors: true  # in case it's not installed

ansible/master-k8s.yaml

@@ -2,3 +2,4 @@
 - import_playbook: k8s-install-deps.yaml
 - import_playbook: k8s-init.yaml
 - import_playbook: install_cni.yaml
+- import_playbook: longhorn-reqs.yaml

init-deploy.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+
+KUBE_USER="kevin"
+KUBE_CP="k8s-control-plane"
+KUBE_W1="k8s-worker1"
+KUBE_W2="k8s-worker2"
+
+ANSIBLE_DIR="ansible"
+
+## Prepare all of the nodes with k8s using the ansible playbooks I prepared
+
+ansible-playbook -i ./"$ANSIBLE_DIR"/inventory.ini ./"$ANSIBLE_DIR"/master-k8s.yaml
+
+## Fetch the configuration from the freshly installed cluster. BEWARE THAT ANY EXISTING CONFIG WILL BE OVERWRITTEN
+
+scp "$KUBE_USER@$KUBE_CP":"$HOME"/.kube/config "$HOME"/.kube/config
+
+## Now join the workers to the cluster
+
+JOIN_TOKEN=$(ssh "$KUBE_USER@$KUBE_CP" "kubeadm token create --print-join-command")
+
+for NODE in "$KUBE_W1" "$KUBE_W2"; do 
+    echo "Joining $NODE"
+    ssh "$KUBE_USER@$NODE" "sudo $JOIN_TOKEN"
+done