Add Ansible playbook for the management cluster (step 1)

VM from Ubuntu cloud image on pv1-1, pinned k3s, cert-manager and cluster-api-operator via k3s HelmChart CRs, CAPI providers as operator CRs, kubeconfig fetch + cluster-inspector kubeconfig. Lint passes (production). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-27 01:02:50 +02:00
parent ac7d5a45b9
commit 66d61b21fc
12 changed files with 440 additions and 0 deletions
@@ -10,3 +10,4 @@ clusterctl.yaml
 # packer / image-builder output
 packer.log
 *.iso
+!ansible/secrets.yml.example
@@ -0,0 +1,7 @@
+[defaults]
+inventory = inventory.yml
+interpreter_python = auto_silent
+
+[ssh_connection]
+# accept-new: the mgmt VM is created by this playbook, so its host key is new on first run
+ssh_args = -o ControlMaster=auto -o ControlPersist=60s -o StrictHostKeyChecking=accept-new
@@ -0,0 +1,38 @@
+# Pinned versions: docs/versions.md is the source of truth, keep in sync.
+k3s_version: v1.35.8+k3s1
+cert_manager_chart_version: v1.21.2
+capi_operator_chart_version: 0.28.0
+capi_version: v1.12.11
+capmox_version: v0.9.1
+ipam_in_cluster_version: v1.1.1
+
+ubuntu_image_url: https://cloud-images.ubuntu.com/noble/20260926/noble-server-cloudimg-amd64.img
+ubuntu_image_sha256: 6a81c37564db9b1ee84e141922625e1d7c5b389b99bb3c572e0243607d5bb4d2
+ubuntu_image_path: /var/lib/vz/template/iso/noble-server-cloudimg-amd64-20260926.img
+
+# Proxmox (token secret lives in secrets.yml)
+proxmox_url: https://192.168.0.8:8006
+proxmox_token_id: capmox@pve!capi
+proxmox_storage: local-lvm
+proxmox_bridge: vmbr0
+proxmox_vlan: 4
+
+# Management VM (VLAN 4). Fill in the empty values before the first run.
+mgmt_vmid: ""
+mgmt_name: lab-mgmt
+mgmt_ip: ""
+mgmt_prefix: 24
+mgmt_gateway: ""
+mgmt_dns: ""
+mgmt_cores: 2
+mgmt_memory_mib: 4096
+mgmt_disk_size: 30G
+mgmt_ssh_public_key: "{{ lookup('file', lookup('env', 'HOME') + '/.ssh/id_ed25519.pub') }}"
+
+# Installed by the cluster-api-operator; CR namespace = install namespace (clusterctl defaults).
+capi_providers:
+  - { kind: CoreProvider, name: cluster-api, namespace: capi-system, version: "{{ capi_version }}" }
+  - { kind: BootstrapProvider, name: kubeadm, namespace: capi-kubeadm-bootstrap-system, version: "{{ capi_version }}" }
+  - { kind: ControlPlaneProvider, name: kubeadm, namespace: capi-kubeadm-control-plane-system, version: "{{ capi_version }}" }
+  - { kind: InfrastructureProvider, name: proxmox, namespace: capmox-system, version: "{{ capmox_version }}", config_secret: capmox-variables }
+  - { kind: IPAMProvider, name: in-cluster, namespace: capi-ipam-in-cluster-system, version: "{{ ipam_in_cluster_version }}" }
@@ -0,0 +1,12 @@
+all:
+  children:
+    pve:
+      hosts:
+        pv1-1:
+          ansible_host: 192.168.0.8
+          ansible_user: root
+    mgmt:
+      hosts:
+        lab-mgmt:
+          ansible_host: "{{ mgmt_ip }}"
+          ansible_user: ubuntu
@@ -0,0 +1,258 @@
+# Management cluster: VM on Proxmox -> k3s -> cert-manager + cluster-api-operator -> CAPI providers.
+# Idempotent: a second run reports changed=0. Run from this directory:
+#   ansible-playbook mgmt.yml [--tags vm|k3s|addons|capi|kubeconfig|inspector]
+
+- name: Management VM on Proxmox
+  hosts: pve
+  gather_facts: false
+  tags: [vm]
+  tasks:
+    - name: Required VM parameters are set in group_vars/all.yml
+      ansible.builtin.assert:
+        that:
+          - mgmt_vmid | string | length > 0
+          - mgmt_ip | length > 0
+          - mgmt_gateway | length > 0
+          - mgmt_dns | length > 0
+
+    - name: Check whether the VM exists
+      ansible.builtin.command: qm status {{ mgmt_vmid }}
+      register: vm_status
+      changed_when: false
+      failed_when: false
+
+    - name: Create the VM from the Ubuntu cloud image
+      when: vm_status.rc != 0
+      block:
+        - name: Download the pinned cloud image
+          ansible.builtin.get_url:
+            url: "{{ ubuntu_image_url }}"
+            dest: "{{ ubuntu_image_path }}"
+            checksum: "sha256:{{ ubuntu_image_sha256 }}"
+            mode: "0644"
+
+        - name: Stage the SSH public key for cloud-init
+          ansible.builtin.copy:
+            content: "{{ mgmt_ssh_public_key }}\n"
+            dest: /root/{{ mgmt_name }}.pub
+            mode: "0600"
+
+        # raw disk on LVM-thin; bridge + separate VLAN tag, never vmbr0.4
+        - name: Create the VM
+          ansible.builtin.command: >-
+            qm create {{ mgmt_vmid }}
+            --name {{ mgmt_name }}
+            --ostype l26
+            --cores {{ mgmt_cores }}
+            --memory {{ mgmt_memory_mib }}
+            --net0 virtio,bridge={{ proxmox_bridge }},tag={{ proxmox_vlan }}
+            --scsihw virtio-scsi-pci
+            --scsi0 {{ proxmox_storage }}:0,import-from={{ ubuntu_image_path }}
+            --ide2 {{ proxmox_storage }}:cloudinit
+            --boot order=scsi0
+            --serial0 socket --vga serial0
+            --agent 1
+            --ciuser ubuntu
+            --sshkeys /root/{{ mgmt_name }}.pub
+            --ipconfig0 ip={{ mgmt_ip }}/{{ mgmt_prefix }},gw={{ mgmt_gateway }}
+            --nameserver {{ mgmt_dns }}
+          changed_when: true
+
+        - name: Grow the disk
+          ansible.builtin.command: qm resize {{ mgmt_vmid }} scsi0 {{ mgmt_disk_size }}
+          changed_when: true
+
+        - name: Start the new VM
+          ansible.builtin.command: qm start {{ mgmt_vmid }}
+          changed_when: true
+
+    - name: Start the VM if it exists but is stopped
+      ansible.builtin.command: qm start {{ mgmt_vmid }}
+      when: vm_status.rc == 0 and 'stopped' in vm_status.stdout
+      changed_when: true
+
+    - name: Wait for SSH on the VM
+      ansible.builtin.wait_for:
+        host: "{{ mgmt_ip }}"
+        port: 22
+        timeout: 300
+      delegate_to: localhost
+
+- name: Install k3s on the management VM
+  hosts: mgmt
+  become: true
+  gather_facts: false
+  tags: [k3s]
+  handlers:
+    - name: Restart k3s
+      ansible.builtin.systemd_service:
+        name: k3s
+        state: restarted
+  tasks:
+    - name: Wait for cloud-init to finish (apt lock, SSH keys)
+      ansible.builtin.command: cloud-init status --wait
+      changed_when: false
+
+    - name: Install qemu-guest-agent
+      ansible.builtin.apt:
+        name: qemu-guest-agent
+        update_cache: true
+        cache_valid_time: 86400
+
+    - name: Start qemu-guest-agent
+      ansible.builtin.systemd_service:
+        name: qemu-guest-agent
+        state: started
+
+    - name: Create the k3s config directory
+      ansible.builtin.file:
+        path: /etc/rancher/k3s
+        state: directory
+        mode: "0755"
+
+    - name: Write the k3s config
+      ansible.builtin.template:
+        src: k3s-config.yaml.j2
+        dest: /etc/rancher/k3s/config.yaml
+        mode: "0644"
+      notify: Restart k3s
+
+    - name: Installed k3s version
+      ansible.builtin.command: k3s --version
+      register: k3s_installed
+      changed_when: false
+      failed_when: false
+
+    - name: Install or upgrade k3s to the pinned version
+      when: k3s_version not in k3s_installed.stdout | default('')
+      block:
+        - name: Download the k3s install script
+          ansible.builtin.get_url:
+            url: https://get.k3s.io
+            dest: /usr/local/bin/k3s-install.sh
+            mode: "0755"
+
+        - name: Run the k3s install script
+          ansible.builtin.command: /usr/local/bin/k3s-install.sh
+          environment:
+            INSTALL_K3S_VERSION: "{{ k3s_version }}"
+          changed_when: true
+
+    - name: Apply pending k3s restart
+      ansible.builtin.meta: flush_handlers
+
+    - name: Wait for the node to be Ready
+      ansible.builtin.command: k3s kubectl wait --for=condition=Ready node --all --timeout=30s
+      register: node_ready
+      until: node_ready.rc == 0
+      retries: 12
+      delay: 10
+      changed_when: false
+
+- name: Install cert-manager and cluster-api-operator (k3s helm-controller)
+  hosts: mgmt
+  become: true
+  gather_facts: false
+  tags: [addons]
+  tasks:
+    - name: Write the cert-manager HelmChart
+      ansible.builtin.template:
+        src: helmchart-cert-manager.yaml.j2
+        dest: /var/lib/rancher/k3s/server/manifests/cert-manager.yaml
+        mode: "0600"
+
+    - name: Wait for cert-manager deployments
+      ansible.builtin.command: k3s kubectl -n cert-manager wait --for=condition=Available deploy --all --timeout=30s
+      register: cm_ready
+      until: cm_ready.rc == 0
+      retries: 30
+      delay: 10
+      changed_when: false
+
+    - name: Write the cluster-api-operator HelmChart
+      ansible.builtin.template:
+        src: helmchart-capi-operator.yaml.j2
+        dest: /var/lib/rancher/k3s/server/manifests/cluster-api-operator.yaml
+        mode: "0600"
+
+    - name: Wait for cluster-api-operator deployment
+      ansible.builtin.command: k3s kubectl -n capi-operator-system wait --for=condition=Available deploy --all --timeout=30s
+      register: op_ready
+      until: op_ready.rc == 0
+      retries: 30
+      delay: 10
+      changed_when: false
+
+    - name: Wait for the operator CRDs
+      ansible.builtin.command: k3s kubectl wait --for=condition=established crd/coreproviders.operator.cluster.x-k8s.io --timeout=60s
+      changed_when: false
+
+- name: CAPI providers
+  hosts: mgmt
+  become: true
+  gather_facts: false
+  tags: [capi]
+  vars_files:
+    - secrets.yml
+  tasks:
+    - name: Manifest directory
+      ansible.builtin.file:
+        path: /opt/lab-mgmt
+        state: directory
+        mode: "0700"
+
+    - name: Provider manifests (namespaces, capmox-variables Secret, provider CRs)
+      ansible.builtin.template:
+        src: capi-providers.yaml.j2
+        dest: /opt/lab-mgmt/capi-providers.yaml
+        mode: "0600"
+      diff: false
+
+    - name: Apply provider manifests
+      ansible.builtin.command: k3s kubectl apply -f /opt/lab-mgmt/capi-providers.yaml
+      register: providers_apply
+      changed_when: "' created' in providers_apply.stdout or ' configured' in providers_apply.stdout"
+
+    - name: Wait for provider deployments
+      ansible.builtin.command: k3s kubectl -n {{ item.namespace }} wait --for=condition=Available deploy --all --timeout=30s
+      loop: "{{ capi_providers }}"
+      loop_control:
+        label: "{{ item.kind }}/{{ item.name }}"
+      register: provider_ready
+      until: provider_ready.rc == 0
+      retries: 40
+      delay: 15
+      changed_when: false
+
+- name: Kubeconfigs on the laptop
+  hosts: mgmt
+  become: true
+  gather_facts: false
+  tags: [kubeconfig]
+  tasks:
+    - name: Read the k3s admin kubeconfig
+      ansible.builtin.slurp:
+        src: /etc/rancher/k3s/k3s.yaml
+      register: k3s_kubeconfig
+
+    - name: Write ~/.kube/lab-mgmt.kubeconfig (server = VM IP, names = lab-mgmt)
+      ansible.builtin.copy:
+        content: >-
+          {{ k3s_kubeconfig.content | b64decode
+             | replace('https://127.0.0.1:6443', 'https://' + mgmt_ip + ':6443')
+             | regex_replace('(?m): default$', ': lab-mgmt') }}
+        dest: "{{ lookup('env', 'HOME') }}/.kube/lab-mgmt.kubeconfig"
+        mode: "0600"
+      delegate_to: localhost
+      become: false
+      diff: false
+
+    - name: Read-only kubeconfig for the cluster-inspector agent
+      ansible.builtin.command: >-
+        {{ playbook_dir }}/../.claude/cluster-inspector/make-kubeconfig.sh lab-mgmt
+        --admin-kubeconfig {{ lookup('env', 'HOME') }}/.kube/lab-mgmt.kubeconfig
+        --apply-rbac
+      delegate_to: localhost
+      become: false
+      changed_when: false
+      tags: [inspector]
@@ -0,0 +1,2 @@
+# cp secrets.yml.example secrets.yml (gitignored) and fill in
+proxmox_token_secret: "<secret of capmox@pve!capi>"
@@ -0,0 +1,33 @@
+{% for p in capi_providers %}
+apiVersion: v1
+kind: Namespace
+metadata:
+  name: {{ p.namespace }}
+---
+{% endfor %}
+# Read by CAPMOX v0.9.1 components; the operator does not watch it, so rotating the token
+# means a new Secret name + updated configSecret reference.
+apiVersion: v1
+kind: Secret
+metadata:
+  name: capmox-variables
+  namespace: capmox-system
+type: Opaque
+stringData:
+  PROXMOX_URL: "{{ proxmox_url }}"
+  PROXMOX_TOKEN: "{{ proxmox_token_id }}"
+  PROXMOX_SECRET: "{{ proxmox_token_secret }}"
+{% for p in capi_providers %}
+---
+apiVersion: operator.cluster.x-k8s.io/v1alpha2
+kind: {{ p.kind }}
+metadata:
+  name: {{ p.name }}
+  namespace: {{ p.namespace }}
+spec:
+  version: {{ p.version }}
+{% if p.config_secret is defined %}
+  configSecret:
+    name: {{ p.config_secret }}
+{% endif %}
+{% endfor %}
@@ -0,0 +1,13 @@
+# Providers are NOT set in chart values (they would become Helm hooks, cluster-api-operator#1052);
+# they are applied as separate CRs from capi-providers.yaml.
+apiVersion: helm.cattle.io/v1
+kind: HelmChart
+metadata:
+  name: cluster-api-operator
+  namespace: kube-system
+spec:
+  repo: https://kubernetes-sigs.github.io/cluster-api-operator
+  chart: cluster-api-operator
+  version: {{ capi_operator_chart_version }}
+  targetNamespace: capi-operator-system
+  createNamespace: true
@@ -0,0 +1,14 @@
+apiVersion: helm.cattle.io/v1
+kind: HelmChart
+metadata:
+  name: cert-manager
+  namespace: kube-system
+spec:
+  repo: https://charts.jetstack.io
+  chart: cert-manager
+  version: {{ cert_manager_chart_version }}
+  targetNamespace: cert-manager
+  createNamespace: true
+  valuesContent: |-
+    crds:
+      enabled: true
@@ -0,0 +1,5 @@
+write-kubeconfig-mode: "0600"
+tls-san:
+  - {{ mgmt_ip }}
+disable:
+  - traefik
@@ -0,0 +1,57 @@
+# Execution log: idempotent management cluster (Ansible + cluster-api-operator)
+
+Pairs with [docs/plans/2026-09-27-0058-mgmt-cluster-ansible.md](../plans/2026-09-27-0058-mgmt-cluster-ansible.md).
+
+## Status
+
+- [x] Step 0 — Initial import commit + plan commit
+- [x] Step 1 — Scaffold `ansible/`, `.gitignore` exception, syntax/lint checks
+- [ ] Step 2 — Run the playbook tag by tag against the lab (owner)
+- [ ] Step 3 — Idempotency check (second full run `changed=0`)
+- [ ] Step 4 — Docs, versions, decisions, changelog
+
+## Step 0 — Initial import + plan
+
+The repo had no commits, so the existing tree went in first as `5fd998c` (after a grep for
+token/secret/key material found only placeholders), then the plan alone as `ac7d5a4`.
+
+## Step 1 — Scaffold `ansible/` (2026-09-27 01:02 CEST)
+
+Wrote `ansible/{ansible.cfg,inventory.yml,group_vars/all.yml,secrets.yml.example,mgmt.yml}` and
+four templates as planned. `ansible/secrets.yml` was already ignored by `*secret*`; added
+`!ansible/secrets.yml.example`.
+
+`ansible-core` is not installed on the laptop, so checks ran through `uvx` (cached, nothing
+installed globally), pinned to ansible-core 2.18 (resolved 2.18.19):
+
+```bash
+cd ansible
+uvx --from 'ansible-core>=2.18,<2.19' ansible-playbook mgmt.yml --syntax-check
+uvx --with 'ansible-core>=2.18,<2.19' ansible-lint --profile production mgmt.yml
+```
+
+First lint run: 6× `name[casing]`, 5× `no-changed-when`. The commands flagged are the
+create/resize/start/install tasks that only run behind a `when:` guard, so they got
+`changed_when: true` rather than a fake condition. Second run: production profile passed.
+
+Rendered all templates with dummy values (`-e mgmt_ip=192.0.2.10`, secret `DUMMY`) through a
+throwaway local playbook: provider manifest parses as 11 documents (5 Namespaces, Secret, 5
+provider CRs), k3s config and kubeconfig rewrite (`server` → VM IP, every `default` name →
+`lab-mgmt`) come out as intended.
+
+Connectivity to the PVE host:
+
+```bash
+uvx --from 'ansible-core>=2.18,<2.19' ansible pve -m ping -o
+# pv1-1 | UNREACHABLE!: ... root@192.168.0.8: Permission denied (publickey,password).
+```
+
+**Worth noting:**
+- The plan's assumption "root SSH from the laptop to `pv1-1` works" does not hold yet; the host
+  answers but rejects the laptop's keys. Needed before Step 2.
+- Condition-based waits: the playbook waits for `Available` Deployments in each provider
+  namespace instead of a condition on the provider CRs, so it doesn't depend on the operator's
+  condition naming (plan left that open).
+- VM params (`mgmt_vmid`, `mgmt_ip`, `mgmt_gateway`, `mgmt_dns`) are empty in `group_vars/all.yml`;
+  Play 1 asserts on them.
+- `clusterctl` is not installed on the laptop either (needed for phase-1 Step 3, not for this playbook).