fix: use import instead of include

.editorconfig

@@ -12,3 +12,7 @@ indent_style = space
 [*.json]
 indent_size = 2
 indent_style = space
+
+[*.{yml,yaml}]
+indent_size = 2
+indent_style = space

.gitignore

@@ -2,8 +2,11 @@
 .*
 terraform.tfstate
 terraform.tfstate.backup
+ansible/lookup_plugins
+ansible/galaxy_roles
 
 !.gitignore
 !.editorconfig
 !.vscode
 !.pre-commit-config.yaml
+!.yamlfmt.yaml

.pre-commit-config.yaml

@@ -9,8 +9,16 @@ repos:
         args:
           - --autofix
       - id: trailing-whitespace
+  - repo: https://github.com/google/yamlfmt
+    rev: v0.21.0
+    hooks:
+      - id: yamlfmt
   - repo: https://github.com/tofuutils/pre-commit-opentofu
     rev: v2.4.2
     hooks:
       - id: tofu_validate
       - id: tofu_fmt
+  - repo: https://github.com/ansible/ansible-lint
+    rev: v26.4.0
+    hooks:
+      - id: ansible-lint

.vscode/extensions.json

@@ -2,6 +2,7 @@
   "recommendations": [
     "EditorConfig.EditorConfig",
     "opentofu.vscode-opentofu",
-    "skellock.just"
+    "skellock.just",
+    "redhat.ansible"
   ]
 }

.yamlfmt.yaml

@@ -0,0 +1,5 @@
+formatter:
+  indent: 2
+  retain_line_breaks_single: true
+  scan_folded_as_literal: true
+  indentless_arrays: false

Justfile

@@ -1,26 +1,39 @@
 #!/usr/bin/env just
 
 opentofu_dir := "infra/"
+ansible_proton_pass_plugin_ref := "4bd0741c347646060ec59c73f8adf7ed8e706cf3"
+ansible_proton_pass_plugin_url := "https://raw.githubusercontent.com/protonpass/proton-pass-ansible-integration/" + ansible_proton_pass_plugin_ref + "/lookup_plugins/proton_pass.py"
 
+init: tofu-init fetch-ansible-deps fetch-ansible-plugins
 up: tofu-apply ansible-run-playbook
 down: tofu-destroy
 
 tofu-init:
     tofu -chdir={{ opentofu_dir }} init -upgrade
 
-tofu-apply: tofu-init
+tofu-apply:
     tofu -chdir={{ opentofu_dir }} apply -auto-approve
 
-tofu-destroy: tofu-init
+tofu-destroy:
     tofu -chdir={{ opentofu_dir }} apply -auto-approve -destroy
 
+tofu-destroy-instances:
+    tofu -chdir={{ opentofu_dir }} apply -auto-approve -destroy -target=module.master -target=module.worker
+
 ansible-run-playbook:
-    ansible-playbook -i ansible/inventory/incus.yaml -c community.general.incus ansible/playbook.yaml
+    ansible-playbook -i ansible/inventories/incus/incus.yaml -c community.general.incus ansible/playbooks/provisioning.yaml -e ansible_user=root
+    ansible-playbook -i ansible/inventories/incus/incus.yaml -c community.general.incus ansible/site.yaml
 
-clean:
+headlamp:
-    rm ansible/hosts.ini || true
+    incus exec master-0 -- cat /etc/kubernetes/admin.conf > .incus-admin.conf
-    rm -r ansible/.ssh/ || true
+    flatpak run io.kinvolk.Headlamp --kubeconfig "$(pwd)/.incus-admin.conf"
-    rm admin.conf || true
 
-get-admin-conf:
+clean: tofu-destroy
-    incus exec master-0 -- cat /etc/kubernetes/admin.conf > k8s-admin.conf
+    git clean -fdx
+
+fetch-ansible-deps:
+    ansible-galaxy install --role-file ansible/requirements.yml --roles-path ansible/galaxy_roles
+
+fetch-ansible-plugins:
+    mkdir -p ansible/lookup_plugins
+    curl -fsSL -o ansible/lookup_plugins/proton_pass.py {{ ansible_proton_pass_plugin_url }}

README.md

@@ -0,0 +1,67 @@
+# Home Infrastructure
+
+An Ansible + OpenTofu project for installing a minimal Kubernetes installation
+to bare-metal and virtual machines.
+
+## Quick Start
+
+Get started with a local Kubernetes cluster which can be used to design and test
+new configuration before pushing them to real machines.
+
+1. Login with Proton Pass and start the SSH daemon.
+
+    ```shell
+    pass-cli login
+    pass-cli ssh-agent daemon start
+    export SSH_AUTH_SOCK="$HOME/.ssh/proton-pass-agent.sock"
+    ```
+
+2. Install the Proton Pass plugin for Ansible
+
+    ```shell
+    just fetch-plugins
+    ```
+
+3. Launch Incus virtual machines
+
+    ```shell
+    just up
+    ```
+
+4. Tear down the Incus virtual machines
+
+    ```shell
+    just down
+    ```
+
+## Setup
+
+```shell
+just fetch-plugins
+```
+
+## Bare-metal provisioning
+
+When creating a set of fresh machines, you must run the provisioning step to create the "Kolibri" automation user.
+
+The Kolibri automation user is used for all other playbooks and has permissions to execute passwordless sudo.
+
+```shell
+ansible-playbook -i ansible/inventories/bare_metal/hosts.yaml ansible/playbooks/provision.yaml -K -e ansible_user=<your_interactive_user>
+```
+
+## Local Testing with Incus
+
+```shell
+just up
+```
+
+```shell
+just headlamp
+```
+
+> The above command runs the Flatpak version of Headlamp which may not have permissions
+> to access the Kubernetes config file copied from the instance.
+>
+> You should run `flatpak override --user --filesystem=$(pwd) io.kinvolk.Headlamp`
+> to ensure that Headlamp can access the directory contents.

ansible.cfg

@@ -0,0 +1,4 @@
+[defaults]
+roles_path = ansible/roles:ansible/galaxy_roles
+lookup_plugins = ansible/lookup_plugins
+callback_result_format = yaml

ansible/inventories/hive/group_vars/all.yaml

@@ -0,0 +1,8 @@
+ansible_python_interpreter: /usr/bin/python3
+ansible_become_exe: sudo.ws
+ansible_user: kolibri
+vip_interface: eno1
+vip_address: 10.0.0.200
+cluster_name: hive
+cluster_pod_subnet: 172.16.0.0/16
+cluster_service_subnet: 10.96.0.0/16

ansible/inventories/hive/hosts.yaml

@@ -0,0 +1,8 @@
+masters:
+  hosts:
+    vineta:
+      ansible_host: vineta.local
+    rotfront:
+      ansible_host: rotfront.local
+    leng:
+      ansible_host: leng.local

ansible/inventories/incus/group_vars/all.yaml

@@ -0,0 +1,7 @@
+ansible_python_interpreter: /usr/bin/python3
+ansible_user: kolibri
+vip_interface: enp5s0
+vip_address: 10.150.0.100
+cluster_name: hive
+cluster_pod_subnet: 172.16.0.0/16
+cluster_service_subnet: 10.96.0.0/16

ansible/inventories/incus/incus.yaml

@@ -0,0 +1,7 @@
+plugin: community.general.incus
+strict: true
+remotes:
+  - local:homelab
+groups:
+  masters: "'master' in inventory_hostname"
+  workers: "'worker' in inventory_hostname"

ansible/inventory/incus.yaml

@@ -1,5 +0,0 @@
----
-plugin: community.general.incus
-strict: true
-remotes:
-  - local:deskpi-cluster

ansible/playbook.yaml

@@ -1,3 +0,0 @@
----
-- name: Provision with Incus
-  ansible.builtin.import_playbook: playbooks/common.yaml

ansible/playbooks/common.yaml

@@ -1,22 +0,0 @@
----
-- name: Bootstrap - Stage 1
-  hosts: all
-  gather_facts: false
-  tasks:
-    - name: Install Python 3
-      ansible.builtin.raw: |
-        if ! command -v python3 >/dev/null 2>&1; then
-          apk add --no-cache python3
-        fi
-      register: apk_result
-      changed_when: "'OK' in apk_result.stdout or 'Installing' in apk_result.stdout"
-
-- name: Bootstrap - Stage 2
-  hosts: all
-  tasks:
-    - name: Install Python 3 Libraries
-      ansible.builtin.package:
-        name: "{{ item }}"
-      loop:
-        - py3-yaml
-        - py3-kubernetes

ansible/playbooks/master-playbook.yaml

@@ -1,154 +0,0 @@
-- name: Primary Master Initialisation
-  hosts: masters[0]
-  tasks:
-    - name: Initialise high-availability Kubernetes cluster
-      ansible.builtin.command: >-
-        kubeadm init
-        --pod-network-cidr="172.16.0.0/16"
-        --control-plane-endpoint="{{ ansible_default_ipv4.address }}:6443"
-        --upload-certs
-      args:
-        creates: /etc/kubernetes/admin.conf
-    - name: Copy Kubernetes Admin Config
-      ansible.builtin.command: "{{ item }}"
-      loop:
-        - mkdir -p $HOME/.kube
-        - cp /etc/kubernetes/admin.conf $HOME/.kube/config
-        - chown $(id -u):$(id -g) $HOME/.kube/config
-      args:
-        creates: $HOME/.kube/config
-
-    - name: Install Pod Network (Calico)
-      ansible.builtin.shell: kubectl apply -f https://raw.githubusercontent.com/projectcalico/calico/v3.32.0/manifests/calico.yaml > pod_network_setup.txt
-      args:
-        chdir: $HOME
-        creates: pod_network_setup.txt
-
-    - name: Add Helm repositories
-      kubernetes.core.helm_repository:
-        name: "{{ item.name }}"
-        repo_url: "{{ item.url }}"
-      loop:
-        - name: metrics-server
-          url: https://kubernetes-sigs.github.io/metrics-server/
-        - name: prometheus-community
-          url: https://prometheus-community.github.io/helm-charts
-        - name: longhorn
-          url: https://charts.longhorn.io
-        - name: traefik
-          url: https://traefik.github.io/charts
-        - name: minecraft-server-charts
-          url: https://itzg.github.io/minecraft-server-charts/
-
-    - name: Install Metrics Server
-      kubernetes.core.helm:
-        release_name: metrics-server
-        release_namespace: kube-system
-        chart_ref: metrics-server/metrics-server
-        chart_version: 3.13.0
-        values:
-          args:
-            - --kubelet-insecure-tls
-
-    - name: Install Longhorn
-      kubernetes.core.helm:
-        release_name: longhorn
-        release_namespace: longhorn-system
-        create_namespace: true
-        chart_ref: longhorn/longhorn
-        chart_version: 1.12.0
-
-    - name: Install Prometheus
-      kubernetes.core.helm:
-        release_name: prometheus
-        release_namespace: monitoring
-        create_namespace: true
-        chart_ref: prometheus-community/prometheus
-        chart_version: 29.10.0
-        state: absent
-
-    - name: Install Traefik
-      kubernetes.core.helm:
-        release_name: traefik
-        release_namespace: default
-        create_namespace: true
-        chart_ref: traefik/traefik
-        chart_version: 40.2.0
-        values:
-          deployment:
-            kind: DaemonSet
-          securityContext:
-            capabilities:
-              drop: [ALL]
-              add: [NET_BIND_SERVICE]
-            readOnlyRootFilesystem: true
-            allowPrivilegeEscalation: false
-          ports:
-            web:
-              port: 80
-              containerPort: 80
-              hostPort: 80
-            websecure:
-              port: 443
-              containerPort: 443
-              hostPort: 443
-
-    - name: Install Minecraft
-      kubernetes.core.helm:
-        release_name: minecraft
-        release_namespace: default
-        chart_ref: minecraft-server-charts/minecraft
-        chart_version: 5.1.3
-        values:
-          minecraftServer:
-            eula: true
-          persistence:
-            dataDir:
-              enabled: true
-
-    - name: Extract Control Plane Decryption Key & Join Token
-      block:
-        - name: Generate Fresh Join Command
-          ansible.builtin.command: kubeadm token create --print-join-command
-          register: join_command_raw
-
-        - name: Upload Certs and Capture Certificate Key
-          shell: kubeadm init phase upload-certs --upload-certs | tail -n 1
-          register: cert_key_raw
-
-        - name: Set facts across playbooks
-          ansible.builtin.set_fact:
-            k8s_join_base: "{{ join_command_raw.stdout }}"
-            k8s_cert_key: "{{ cert_key_raw.stdout }}"
-          delegate_to: localhost
-          delegate_facts: true
-
-
-- name: Join Secondary Control Planes
-  hosts: masters:!masters[0] # <--- Targets masters 1 and 2
-  tasks:
-    - name: Join Cluster as Control Plane Node
-      ansible.builtin.shell: "{{ hostvars['localhost']['k8s_join_base'] }} --control-plane --certificate-key {{ hostvars['localhost']['k8s_cert_key'] }}"
-      args:
-        creates: /etc/kubernetes/admin.conf
-
-    - name: Copy Kubernetes Admin Config
-      ansible.builtin.shell: "{{ item }}"
-      loop:
-        - mkdir -p $HOME/.kube
-        - cp /etc/kubernetes/admin.conf $HOME/.kube/config
-        - chown $(id -u):$(id -g) $HOME/.kube/config
-      args:
-        creates: $HOME/.kube/config
-
-- name: Remove control plane taint
-  hosts: masters[0]
-  tasks:
-    - name: Remove control plane NoSchedule taint from all nodes
-      kubernetes.core.k8s_taint:
-        state: absent
-        name: "{{ item }}"
-        taints:
-          - key: node-role.kubernetes.io/control-plane
-            effect: NoSchedule
-      loop: "{{ groups['masters'] }}"

ansible/playbooks/performance.yaml

@@ -0,0 +1,7 @@
+---
+- name: Configure performance tuning
+  hosts: all
+  become: true
+  roles:
+    - giovtorres.tuned
+    - irqbalance

ansible/playbooks/provisioning.yaml

@@ -0,0 +1,25 @@
+- name: Create Kolibri automation user
+  hosts: all
+  become: true
+  tasks:
+    - name: Ensure Kolibri user exists
+      ansible.builtin.user:
+        name: kolibri
+        comment: Hummingbird
+        create_home: true
+        groups:
+          - sudo
+        shell: /bin/bash
+        system: true
+
+    - name: Ensure authorised keys for Kolibri
+      ansible.posix.authorized_key:
+        user: kolibri
+        key: "{{ lookup('proton_pass', vault_name='Home Lab', item_title='Kolibri Automation', field='Public key') }}"
+
+    - name: Allow passwordless sudo for Kolibri
+      ansible.builtin.copy:
+        content: "kolibri ALL=(ALL) NOPASSWD: ALL\n"
+        dest: /etc/sudoers.d/kolibri
+        mode: "0440"
+        validate: visudo -cf %s

ansible/playbooks/reset.yaml

@@ -0,0 +1,13 @@
+- name: Reset nodes
+  hosts: all
+  become: true
+  tasks:
+    - name: Reset kubeadm
+      ansible.builtin.command: kubeadm reset -f
+      args:
+        removes: /etc/kubernetes/*.conf
+
+    - name: Clear CNI configuration
+      ansible.builtin.file:
+        path: /etc/cni/net.d
+        state: absent

ansible/playbooks/setup-playbook.yaml

@@ -1,87 +0,0 @@
-- name: Initialise Master and Worker Nodes
-  hosts: all
-  tasks:
-    - name: Enable IPv4 forwarding
-      ansible.posix.sysctl:
-        name: net.ipv4.ip_forward
-        value: 1
-        sysctl_set: true
-
-    - name: Enable Kernel Modules
-      community.general.modprobe:
-        name: "{{ item }}"
-        persistent: present
-      loop:
-        - br_netfilter
-        - dm-crypt
-
-    - name: Ensure /sys mount is set to shared for container runtimes
-      block:
-        - name: Remount /sys as shared immediately
-          ansible.builtin.command: mount --make-rshared {{ item }}
-          loop:
-            - /
-            - /sys
-            - /run
-          changed_when: false
-
-        - name: Create Alpine local.d script for permanent shared mount
-          ansible.builtin.copy:
-            dest: /etc/local.d/mount-shared.start
-            content: |
-              #!/bin/sh
-              mount --make-rshared /
-              mount --make-rshared /sys
-              mount --make-rshared /run
-            owner: root
-            group: root
-            mode: '0755'
-
-        - name: Ensure Alpine local service is enabled on boot
-          ansible.builtin.service:
-            name: local
-            enabled: true
-            runlevel: default
-
-    - name: Install tools
-      ansible.builtin.package:
-        name: "{{ item }}"
-        state: present
-      loop:
-        - htop
-
-    - name: Install kubectl, kubelet, kubeadm and containerd
-      ansible.builtin.package:
-        name: "{{ item }}"
-        state: present
-      loop:
-        - kubectl
-        - kubeadm
-        - kubelet
-        - containerd
-        - cni-plugins
-        - helm
-        - open-iscsi
-
-    - name: Start kubelet and containerd
-      ansible.builtin.service:
-        name: "{{ item }}"
-        enabled: true
-        state: started
-      loop:
-        - kubelet
-        - containerd
-        - iscsid
-
-    - name: Add /opt/cni/bin to containerd CNI binary directories
-      ansible.builtin.replace:
-        path: /etc/containerd/config.toml
-        regexp: "bin_dirs = \\['/usr/libexec/cni'\\]"
-        replace: "bin_dirs = ['/opt/cni/bin', '/usr/libexec/cni']"
-      notify: Restart containerd
-
-  handlers:
-    - name: Restart containerd
-      ansible.builtin.service:
-        name: containerd
-        state: restarted

ansible/playbooks/worker-playbook.yaml

@@ -1,24 +0,0 @@
-- name: Extract Worker Join Token from Primary Master
-  hosts: masters[0] # <--- Targets only the first master safely
-  gather_facts: false
-  tasks:
-    - name: Generate worker join command
-      ansible.builtin.command: kubeadm token create --print-join-command
-      register: join_command_raw
-      changed_when: false # Reading/generating a token text string changes no host state
-
-    - name: Save join command globally
-      ansible.builtin.set_fact:
-        k8s_worker_join: "{{ join_command_raw.stdout }}"
-      delegate_to: localhost
-      delegate_facts: true
-
-- name: Join Worker Nodes to Cluster
-  hosts: workers
-  gather_facts: false
-  tasks:
-    - name: Join cluster
-      ansible.builtin.shell: "{{ hostvars['localhost']['k8s_worker_join'] }} >> node_joined.txt"
-      args:
-        chdir: $HOME
-        creates: node_joined.txt

ansible/requirements.yml

@@ -0,0 +1,6 @@
+---
+collections:
+  - ansible.posix
+roles:
+  - name: giovtorres.tuned
+    version: 2.0.2

ansible/roles/common/files/ssh.service

@@ -0,0 +1,7 @@
+<service-group>
+  <name replace-wildcards="yes">%h</name>
+  <service>
+    <type>_ssh._tcp</type>
+    <port>22</port>
+  </service>
+</service-group>

ansible/roles/common/handlers/main.yaml

@@ -0,0 +1,4 @@
+- name: Restart Avahi Daemon
+  ansible.builtin.service:
+    name: avahi-daemon
+    state: restarted

ansible/roles/common/tasks/avahi.yaml

@@ -0,0 +1,19 @@
+- name: Install Avahi Daemon
+  ansible.builtin.apt:
+    pkg:
+      - avahi-daemon
+    state: present
+
+- name: Enable Avahi Daemon Service
+  ansible.builtin.service:
+    name: avahi-daemon
+    state: started
+    enabled: true
+
+- name: Copy Avahi SSH service configuration
+  ansible.builtin.copy:
+    src: ssh.service
+    dest: /etc/avahi/services/ssh.service
+    mode: "0644"
+  notify:
+    - Restart Avahi Daemon

ansible/roles/common/tasks/main.yaml

@@ -0,0 +1,8 @@
+- name: Install required packages
+  ansible.builtin.import_tasks: packages.yaml
+
+- name: Setup networking configuration
+  ansible.builtin.import_tasks: network.yaml
+
+- name: Setup Avahi Daemon
+  ansible.builtin.import_tasks: avahi.yaml

ansible/roles/common/tasks/network.yaml

@@ -0,0 +1,5 @@
+- name: Enable IPv4 Forwarding
+  ansible.posix.sysctl:
+    name: net.ipv4.ip_forward
+    value: "1"
+    sysctl_set: true

ansible/roles/common/tasks/packages.yaml

@@ -0,0 +1,6 @@
+- name: Install Ansible Dependencies
+  ansible.builtin.apt:
+    pkg:
+      - python3-debian
+      - python3-kubernetes
+    state: present

ansible/roles/irqbalance/tasks/main.yaml

@@ -0,0 +1,11 @@
+---
+- name: Install irqbalance
+  ansible.builtin.package:
+    name: irqbalance
+    state: present
+
+- name: Ensure irqbalance service
+  ansible.builtin.service:
+    name: irqbalance
+    state: started
+    enabled: true

ansible/roles/kubernetes_common/tasks/main.yaml

@@ -0,0 +1,6 @@
+- name: Install Kubernetes APT repository
+  ansible.builtin.import_tasks: repo.yaml
+- name: Install Kubernetes packages
+  ansible.builtin.import_tasks: packages.yaml
+- name: Ensure services running
+  ansible.builtin.import_tasks: services.yaml

ansible/roles/kubernetes_common/tasks/packages.yaml

@@ -0,0 +1,9 @@
+- name: Install Kubernetes packages
+  ansible.builtin.apt:
+    pkg:
+      - kubeadm
+      - kubelet
+      - containerd
+      - kubectl
+      - helm
+    update_cache: true

ansible/roles/kubernetes_common/tasks/repo.yaml

@@ -0,0 +1,18 @@
+- name: Add Kubernetes APT repository
+  ansible.builtin.deb822_repository:
+    name: kubernetes
+    types: deb
+    uris: https://pkgs.k8s.io/core:/stable:/v1.36/deb/
+    suites: /
+    signed_by: https://pkgs.k8s.io/core:/stable:/v1.36/deb/Release.key
+    state: present
+
+- name: Add Helm APT repository
+  ansible.builtin.deb822_repository:
+    name: helm
+    types: deb
+    uris: https://packages.buildkite.com/helm-linux/helm-debian/any/
+    suites: any
+    components: main
+    signed_by: https://packages.buildkite.com/helm-linux/helm-debian/gpgkey
+    state: present

ansible/roles/kubernetes_common/tasks/services.yaml

@@ -0,0 +1,5 @@
+- name: Ensure containerd is running
+  ansible.builtin.service:
+    name: containerd
+    state: started
+    enabled: true

ansible/roles/kubernetes_control_plane/tasks/copy-config.yaml

@@ -0,0 +1,18 @@
+- name: Setup Kubernetes admin config for root user
+  become: true
+  block:
+    - name: Ensure .kube directory
+      ansible.builtin.file:
+        path: "{{ ansible_facts['env']['HOME'] }}/.kube"
+        state: directory
+        owner: "{{ ansible_facts['user_id'] }}"
+        group: "{{ ansible_facts['user_id'] }}"
+        mode: "0755"
+    - name: Copy admin configuration
+      ansible.builtin.copy:
+        src: /etc/kubernetes/admin.conf
+        dest: "{{ ansible_facts['env']['HOME'] }}/.kube/config"
+        remote_src: true
+        owner: "{{ ansible_facts['user_id'] }}"
+        group: "{{ ansible_facts['user_id'] }}"
+        mode: "0600"

ansible/roles/kubernetes_control_plane/tasks/init-cluster.yaml

@@ -0,0 +1,21 @@
+- name: Check if cluster is initialised
+  ansible.builtin.stat:
+    path: /etc/kubernetes/admin.conf
+  register: kubernetes_control_plane_admin_conf
+  run_once: true
+
+- name: Initialise Kubernetes cluster
+  when: not kubernetes_control_plane_admin_conf.stat.exists
+  block:
+    - name: Copy kubeadm config
+      ansible.builtin.template:
+        src: kubeadm-config.yaml.j2
+        dest: /tmp/kubeadm-config.yaml
+        mode: "0600"
+      run_once: true
+
+    - name: Initialise Kubernetes with kubeadm
+      ansible.builtin.command: kubeadm init --config /tmp/kubeadm-config.yaml
+      args:
+        creates: /etc/kubernetes/admin.conf
+      run_once: true

ansible/roles/kubernetes_control_plane/tasks/join-cluster.yaml

@@ -0,0 +1,20 @@
+- name: Generate join command
+  ansible.builtin.command: kubeadm token create --print-join-command --kubeconfig /etc/kubernetes/admin.conf
+  register: kubernetes_control_plane_join_command
+  run_once: true
+  changed_when: false
+
+- name: Get certificate key
+  ansible.builtin.command: kubeadm init phase upload-certs --upload-certs --kubeconfig /etc/kubernetes/admin.conf
+  register: kubernetes_control_plane_certificate_key
+  run_once: true
+  changed_when: false
+
+- name: Join additional control planes
+  ansible.builtin.command: >-
+    {{ kubernetes_control_plane_join_command.stdout }}
+    --control-plane
+    --certificate-key
+    {{ kubernetes_control_plane_certificate_key.stdout_lines[-1] }}
+  args:
+    creates: /etc/kubernetes/admin.conf

ansible/roles/kubernetes_control_plane/tasks/kube-vip.yaml

@@ -0,0 +1,5 @@
+- name: Ensure kube-vip static pod manifest exists
+  ansible.builtin.template:
+    src: kube-vip.yaml.j2
+    dest: /etc/kubernetes/manifests/kube-vip.yaml
+    mode: "0600"

ansible/roles/kubernetes_control_plane/tasks/main.yaml

@@ -0,0 +1,14 @@
+- name: Configure kube-vip static pod manifest
+  ansible.builtin.import_tasks: kube-vip.yaml
+
+- name: Initialise Kubernetes cluster
+  ansible.builtin.import_tasks: init-cluster.yaml
+
+- name: Join nodes as control planes
+  ansible.builtin.import_tasks: join-cluster.yaml
+
+- name: Copy Kubernetes admin configuration on all nodes
+  ansible.builtin.import_tasks: copy-config.yaml
+
+- name: Remove control plane scheduling restrictions
+  ansible.builtin.import_tasks: untaint.yaml

ansible/roles/kubernetes_control_plane/tasks/untaint.yaml

@@ -0,0 +1,7 @@
+- name: Remove control plane NoSchedule taint
+  kubernetes.core.k8s_taint:
+    state: absent
+    name: "{{ ansible_facts['hostname'] }}"
+    taints:
+      - key: node-role.kubernetes.io/control-plane
+        effect: NoSchedule

ansible/roles/kubernetes_control_plane/templates/kube-vip.yaml.j2

@@ -0,0 +1,70 @@
+apiVersion: v1
+kind: Pod
+metadata:
+  name: kube-vip
+  namespace: kube-system
+spec:
+  containers:
+  - args:
+    - manager
+    env:
+    - name: vip_arp
+      value: "true"
+    - name: port
+      value: "6443"
+    - name: vip_nodename
+      valueFrom:
+        fieldRef:
+          fieldPath: spec.nodeName
+    - name: vip_interface
+      value: {{ vip_interface }}
+    - name: vip_subnet
+      value: "32"
+    - name: dns_mode
+      value: first
+    - name: dhcp_mode
+      value: ipv4
+    - name: cp_enable
+      value: "true"
+    - name: cp_namespace
+      value: kube-system
+    - name: svc_enable
+      value: "false"
+    - name: vip_leaderelection
+      value: "true"
+    - name: vip_leasename
+      value: plndr-cp-lock
+    - name: vip_leaseduration
+      value: "15"
+    - name: vip_renewdeadline
+      value: "10"
+    - name: vip_retryperiod
+      value: "2"
+    - name: address
+      value: {{ vip_address }}
+    - name: prometheus_server
+      value: :2112
+    image: ghcr.io/kube-vip/kube-vip:v1.2.1
+    imagePullPolicy: IfNotPresent
+    name: kube-vip
+    resources: {}
+    securityContext:
+      capabilities:
+        add:
+        - NET_ADMIN
+        - NET_RAW
+        drop:
+        - ALL
+    volumeMounts:
+    - mountPath: /etc/kubernetes/admin.conf
+      name: kubeconfig
+  hostAliases:
+  - hostnames:
+    - kubernetes
+    ip: 127.0.0.1
+  hostNetwork: true
+  volumes:
+  - hostPath:
+      path: /etc/kubernetes/admin.conf
+    name: kubeconfig
+status: {}

ansible/roles/kubernetes_control_plane/templates/kubeadm-config.yaml.j2

@@ -0,0 +1,18 @@
+# Configuration reference can be found at https://kubernetes.io/docs/reference/config-api/kubeadm-config.v1beta4/
+apiVersion: kubeadm.k8s.io/v1beta4
+kind: InitConfiguration
+---
+apiVersion: kubeadm.k8s.io/v1beta4
+kind: ClusterConfiguration
+clusterName: "{{ cluster_name }}"
+controlPlaneEndpoint: "{{ vip_address }}"
+networking:
+  podSubnet: "{{ cluster_pod_subnet }}"
+  serviceSubnet: "{{ cluster_service_subnet }}"
+---
+apiVersion: kubelet.config.k8s.io/v1beta1
+kind: KubeletConfiguration
+cgroupDriver: systemd
+failSwapOn: false
+memorySwap:
+  swapBehavior: LimitedSwap

ansible/site.yaml

@@ -0,0 +1,15 @@
+- name: Apply power and performance settings
+  ansible.builtin.import_playbook: playbooks/performance.yaml
+
+- name: Baseline configuration
+  hosts: all
+  become: true
+  roles:
+    - common
+    - kubernetes_common
+
+- name: Initialise Kubernetes Cluster
+  hosts: masters
+  become: true
+  roles:
+    - kubernetes_control_plane

infra/bootstrap.sh

@@ -1,32 +0,0 @@
-#!/usr/bin/env sh
-set -e
-
-err() { echo "$@" >&2; exit 1; }
-log() { echo "$@" >&2; }
-
-[ "$(id -u)" -ne 0 ] && err "You must run this script as root"
-
-# Install minimal packages for provisioning
-PACKAGES="dropbear openssh-sftp-server python3 py3-yaml py3-kubernetes cloud-utils-growpart e2fsprogs-extra lsblk"
-apk update && apk upgrade
-apk add ${PACKAGES}
-
-# Enable and start services
-enable_and_start() {
-    service="$1"
-    rc-update add "${service}" default
-    service "${service}" start
-}
-
-enable_and_start dropbear
-
-# increase root partition size
-root_dev=$(readlink -f /dev/root)
-disk=$(lsblk -no pkname "$root_dev")
-part_num=$(echo "$root_dev" | sed 's|.*[^0-9]||')
-
-echo "Growing /dev/${disk}${part_num}..."
-
-# Grow partition and resize
-growpart "/dev/$disk" "$part_num" || err "Partition growth failed"
-resize2fs "$root_dev" || err "Filesystem resize failed"

infra/main.tofu

@@ -3,12 +3,10 @@
 # ------
 
 locals {
+  image          = "ubuntu/26.04"
   root_disk_size = "16GiB"
   ext_disk_size  = "16GiB"
   network_cidr   = "10.150.0.1/24"
-
-  master_count = 2
-  worker_count = 0
 }
 
 # ---------
@@ -16,49 +14,23 @@ locals {
 # ---------
 
 resource "incus_project" "this" {
-  name        = "deskpi-cluster"
+  name        = "homelab"
-  description = "The emulated DeskPi Super6C stack."
+  description = "An emulated home infrastructure stack."
 }
 
-resource "incus_image" "alpine" {
+resource "incus_image" "this" {
   project = incus_project.this.name
 
   source_image = {
     remote = "images"
-    name   = "alpine/edge"
+    name   = local.image
     type   = "virtual-machine"
   }
 }
 
-resource "incus_profile" "this" {
-  name        = "Default"
-  description = "The default profile applied to both masters and workers."
-  project     = incus_project.this.name
-
-  depends_on = [incus_network.this]
-
-  config = {
-    "limits.cpu"          = var.cpus
-    "limits.memory"       = var.memory
-    "boot.autostart"      = false
-    "security.secureboot" = false
-  }
-
-  device {
-    name = "root"
-    type = "disk"
-
-    properties = {
-      "pool" = "default"
-      "path" = "/"
-      "size" = local.root_disk_size
-    }
-  }
-}
-
 resource "incus_network" "this" {
-  name        = "deskpi-cluster"
+  name        = "homelab"
-  description = "An emulated network for the DeskPi Super6C cluster project."
+  description = "An emulated home infrastructure stack."
   project     = incus_project.this.name
   type        = "bridge"
 
@@ -68,93 +40,32 @@ resource "incus_network" "this" {
   }
 }
 
-resource "incus_instance" "master" {
+module "master" {
-  count = local.master_count
+  source = "./modules/incus_vm"
+  count  = var.master_count
 
-  name        = "master-${count.index}"
+  name    = "master-${count.index}"
-  description = "Kubernetes Master ${count.index}"
+  project = incus_project.this.name
-  project     = incus_project.this.name
+  image   = incus_image.this.fingerprint
+  cpu     = 2
+  memory  = "2GiB"
 
-  type     = "virtual-machine"
+  network      = incus_network.this.name
-  image    = incus_image.alpine.fingerprint
+  ipv4_address = cidrhost(local.network_cidr, 2 + (2 * count.index))
-  profiles = [incus_profile.this.name]
-
-  wait_for {
-    type = "agent"
-  }
-
-  device {
-    name = "eth0"
-    type = "nic"
-
-    properties = {
-      "network"      = incus_network.this.name
-      "ipv4.address" = cidrhost(local.network_cidr, 2 + (2 * count.index))
-    }
-  }
-
-  device {
-    name = "ext"
-    type = "disk"
-
-    properties = {
-      "pool"   = "default"
-      "source" = incus_storage_volume.master[count.index].name
-    }
-  }
 }
 
-resource "incus_storage_volume" "master" {
+module "worker" {
-  count = local.master_count
+  source = "./modules/incus_vm"
+  count  = var.worker_count
 
-  name        = "master-ext-${count.index}"
+  name    = "worker-${count.index}"
-  description = "External drive for master ${count.index}"
+  project = incus_project.this.name
-  pool        = "default"
+  image   = incus_image.this.fingerprint
-  project     = incus_project.this.name
+  cpu     = 2
+  memory  = "2GiB"
 
-  content_type = "block"
+  network      = incus_network.this.name
-
+  ipv4_address = cidrhost(local.network_cidr, 3 + (2 * count.index))
-  config = {
-    "size" = local.ext_disk_size
-  }
-}
-
-resource "incus_instance" "worker" {
-  count = local.worker_count
-
-  name        = "worker-${count.index}"
-  description = "Kubernetes Worker ${count.index}"
-  project     = incus_project.this.name
-
-  type     = "virtual-machine"
-  image    = incus_image.alpine.fingerprint
-  profiles = [incus_profile.this.name]
-
-  wait_for {
-    type = "agent"
-  }
-
-  device {
-    name = "eth0"
-    type = "nic"
-
-    properties = {
-      "network"      = incus_network.this.name
-      "ipv4.address" = cidrhost(local.network_cidr, 3 + (2 * count.index))
-    }
-  }
-}
-
-# -------
-# Outputs
-# -------
-
-output "master_addresses" {
-  value = incus_instance.master[*].ipv4_address
-}
-
-output "worker_addresses" {
-  value = incus_instance.worker[*].ipv4_address
 }
 
 # ---------
@@ -162,12 +73,12 @@ output "worker_addresses" {
 # ---------
 
 terraform {
-  required_version = "~> 1"
+  required_version = ">= 1.11.5"
 
   required_providers {
     incus = {
       source  = "lxc/incus"
-      version = "~> 1"
+      version = "~> 1.1"
     }
   }
 }

infra/modules/incus_vm/main.tofu

@@ -0,0 +1,79 @@
+terraform {
+  required_providers {
+    incus = {
+      source  = "lxc/incus"
+      version = ">= 1.1.1"
+    }
+  }
+}
+
+resource "incus_instance" "this" {
+  name        = var.name
+  description = var.description
+  project     = var.project
+
+  type     = "virtual-machine"
+  image    = var.image
+  profiles = concat([incus_profile.this.name], var.profiles)
+
+  wait_for {
+    type = "agent"
+  }
+
+  device {
+    name = "eth0"
+    type = "nic"
+
+    properties = {
+      "network"      = var.network
+      "ipv4.address" = var.ipv4_address
+    }
+  }
+
+  device {
+    name = "ext"
+    type = "disk"
+
+    properties = {
+      "pool"   = "default"
+      "source" = incus_storage_volume.this.name
+    }
+  }
+}
+
+resource "incus_storage_volume" "this" {
+  name        = "${var.name}-ext"
+  description = "External disk for ${var.name}"
+  pool        = "default"
+  project     = var.project
+
+  content_type = "block"
+
+  config = {
+    "size" = "16GiB"
+  }
+}
+
+resource "incus_profile" "this" {
+  name        = var.name
+  description = "Default profile for ${var.name}"
+  project     = var.project
+
+  config = {
+    "limits.cpu"          = var.cpu
+    "limits.memory"       = var.memory
+    "boot.autostart"      = false
+    "security.secureboot" = true
+  }
+
+  device {
+    name = "root"
+    type = "disk"
+
+    properties = {
+      "pool" = "default"
+      "path" = "/"
+      "size" = "16GiB"
+    }
+  }
+}

infra/modules/incus_vm/variables.tofu

@@ -0,0 +1,55 @@
+variable "name" {
+  type        = string
+  description = "The name for this instance."
+  nullable    = false
+}
+
+variable "description" {
+  type        = string
+  description = "The description for this instance."
+  nullable    = false
+  default     = ""
+}
+
+variable "project" {
+  type        = string
+  description = "The project this instance belongs to."
+  nullable    = true
+}
+
+variable "image" {
+  type        = string
+  description = "The image for this instance."
+  nullable    = false
+}
+
+variable "profiles" {
+  type        = list(string)
+  description = "The profiles for this instance."
+  nullable    = false
+  default     = []
+}
+
+variable "network" {
+  type        = string
+  description = "The network for this instance."
+  nullable    = false
+}
+
+variable "ipv4_address" {
+  type        = string
+  description = "The IPv4 address for this instance."
+  nullable    = false
+}
+
+variable "cpu" {
+  type        = number
+  description = "The number of CPUs for this instance."
+  nullable    = false
+}
+
+variable "memory" {
+  type        = string
+  description = "The amount of memory allocated for this instance."
+  nullable    = false
+}

infra/variables.tofu

@@ -1,3 +1,17 @@
+variable "master_count" {
+  description = "The number of master nodes to provision."
+  type        = number
+  default     = 1
+  nullable    = false
+}
+
+variable "worker_count" {
+  description = "The number of worker nodes to provision."
+  type        = number
+  default     = 0
+  nullable    = false
+}
+
 variable "cpus" {
   description = "The number of CPU cores allocated to each virtual machine."
   type        = number