Fix/eliminate default filter 11822

docs/getting_started/getting-started.md

@@ -6,6 +6,9 @@ Install Ansible according to [Ansible installation guide](/docs/ansible/ansible.
 
 ## Building your own inventory
 
+For a lightweight approach using only the files you need, see the [Minimal Inventory
+guide](minimal-inventory.md).
+
 Ansible inventory can be stored in 3 formats: YAML, JSON, or INI-like. See the
 [example inventory](/inventory/sample/inventory.ini)
 and [Ansible documentation on building your inventory](https://docs.ansible.com/ansible/latest/inventory_guide/intro_inventory.html),

docs/getting_started/minimal-inventory.md

@@ -0,0 +1,244 @@
+# Minimal Inventory Guide
+
+This document describes best practices for managing Kubespray custom configurations
+in a separate git repository with minimal files.
+
+## Why a Minimal Inventory
+
+Many tutorials recommend copying the entire `inventory/sample` directory to
+`inventory/myclustername` and then modifying options directly in those files.
+While this works well for getting started, it results in:
+
+- A large number of files in your custom git repository
+- Difficulty distinguishing which settings differ from Kubespray defaults
+- Increased maintenance burden when Kubespray updates sample files
+
+A better approach is to keep only the files that contain your custom
+configuration and reference Kubespray as a git submodule. This makes it easy
+to track what you have customized and to update Kubespray independently.
+
+## Minimal Inventory Structure
+
+For most clusters, you only need these two files:
+
+```
+your-repo/
+├── inventory/
+│   └── mycluster/
+│       ├── hosts.yaml          # Cluster topology (nodes and groups)
+│       └── group_vars/
+│           └── all.yaml        # Custom configuration overrides
+├── kubespray                   # Submodule pointing to Kubespray
+└── .gitmodules
+```
+
+## Hosts File
+
+Create `inventory/mycluster/hosts.yaml` to define your cluster topology.
+
+### Single-node cluster example
+
+```yaml
+all:
+  hosts:
+    node1:
+      ansible_host: 192.168.1.10
+      # Internal IP for the node (optional)
+      # ip: 10.0.0.2
+  children:
+    kube_control_plane:
+      hosts:
+        node1:
+    kube_node:
+      hosts:
+        node1:
+    etcd:
+      hosts:
+        node1:
+    k8s_cluster:
+      children:
+        kube_control_plane:
+        kube_node:
+```
+
+### Multi-node HA cluster example
+
+```yaml
+all:
+  vars:
+    ansible_user: deploy
+    # ansible_ssh_private_key_file: ~/.ssh/id_rsa
+  hosts:
+    controller1:
+      ansible_host: 192.168.1.10
+    controller2:
+      ansible_host: 192.168.1.11
+    controller3:
+      ansible_host: 192.168.1.12
+    worker1:
+      ansible_host: 192.168.1.20
+    worker2:
+      ansible_host: 192.168.1.21
+    worker3:
+      ansible_host: 192.168.1.22
+  children:
+    kube_control_plane:
+      hosts:
+        controller1:
+        controller2:
+        controller3:
+    kube_node:
+      hosts:
+        worker1:
+        worker2:
+        worker3:
+    etcd:
+      hosts:
+        controller1:
+        controller2:
+        controller3:
+    k8s_cluster:
+      children:
+        kube_control_plane:
+        kube_node:
+```
+
+> **Note:** `etcd` is defined as a child of `kube_control_plane` for a
+"stacked" etcd topology. For external etcd, create a separate `[etcd]` group
+with its own hosts.
+
+## Group Variables File
+
+Create `inventory/mycluster/group_vars/all.yaml` with only the settings that
+differ from Kubespray defaults:
+
+```yaml
+# Cluster identity
+cluster_name: mycluster.example.com
+
+# SSL certificate SAN entries
+supplementary_addresses_in_ssl_keys:
+  - controller1.example.com
+  - controller2.example.com
+  - controller3.example.com
+
+# Enable optional components
+helm_enabled: true
+metrics_server_enabled: true
+metrics_server_kubelet_insecure_tls: true
+krew_enabled: true
+```
+
+You only need to list variables that you want to override. Kubespray has
+sensible defaults for most settings. Check
+`roles/kubespray_defaults/defaults/main/` for the complete list of default
+values.
+
+## Setting Up Kubespray as a Git Submodule
+
+This workflow keeps Kubespray as a dependency rather than copying all files:
+
+1. **Initialize your repository:**
+
+   ```bash
+   git init
+   git remote add origin <your-repo-url>
+   ```
+
+2. **Add Kubespray as a submodule:**
+
+   ```bash
+   git submodule add https://github.com/kubernetes-sigs/kubespray.git kubespray
+   ```
+
+3. **Create your inventory structure:**
+
+   ```bash
+   mkdir -p inventory/mycluster/group_vars
+   ```
+
+4. **Create `hosts.yaml` and `group_vars/all.yaml`** as described above.
+
+5. **Commit your changes:**
+
+   ```bash
+   git add .gitmodules kubespray inventory/
+   git commit -m "Add minimal Kubespray inventory"
+   ```
+
+## Running Kubespray
+
+Navigate to the Kubespray directory and run the playbook:
+
+```bash
+cd kubespray
+ansible-playbook -i ../inventory/mycluster cluster.yml -b -v
+  --private-key=~/.ssh/id_rsa
+```
+
+Or set the `ANSIBLE_INVENTORY` environment variable:
+
+```bash
+export ANSIBLE_INVENTORY=../inventory/mycluster
+cd kubespray
+ansible-playbook cluster.yml -b -v --private-key=~/.ssh/id_rsa
+```
+
+## Organizing Group-Specific Variables
+
+As your configuration grows, you may want to organize variables by group:
+
+```
+inventory/mycluster/
+├── hosts.yaml
+└── group_vars/
+    ├── all.yaml                    # Global settings
+    ├── kube_control_plane.yaml     # Control plane only
+    ├── kube_node.yaml              # Worker nodes only
+    └── etcd.yaml                   # etcd nodes only
+```
+
+Example `group_vars/kube_control_plane.yaml`:
+
+```yaml
+supplementary_addresses_in_ssl_keys:
+  - controller1.example.com
+  - controller2.example.com
+  - controller3.example.com
+```
+
+This approach keeps your configuration organized and makes it clear which
+settings apply to which nodes.
+
+## Using INI Format Instead of YAML
+
+If you prefer the INI format, you can use `hosts.ini`:
+
+```ini
+[kube_control_plane]
+controller1 ansible_host=192.168.1.10
+controller2 ansible_host=192.168.1.11
+controller3 ansible_host=192.168.1.12
+
+[kube_node]
+worker1 ansible_host=192.168.1.20
+worker2 ansible_host=192.168.1.21
+worker3 ansible_host=192.168.1.22
+
+[etcd:children]
+kube_control_plane
+
+[k8s_cluster:children]
+kube_control_plane
+kube_node
+```
+
+## Migrating from inventory/sample
+
+If you have an existing cluster based on `inventory/sample`, you can migrate
+to a minimal inventory:
+
+1. Create a new inventory directory
+2. Copy your `hosts.yaml` (or create a new one from the examples above)
+3. Review your existing `group_vars` and copy only the non-default settings
+4. Test with the new inventory before removing the old files

roles/kubernetes/kubeadm_common/defaults/main.yml

@@ -1,6 +1,7 @@
 ---
 kubeadm_patches_dir: "{{ kube_config_dir }}/patches"
 kubeadm_patches: []
+
 # See https://kubernetes.io/docs/setup/production-environment/tools/kubeadm/control-plane-flags/#patches
 # Correspondance with this link
 # patchtype = type

roles/kubernetes/node-label/defaults/main.yml

@@ -0,0 +1,7 @@
+# Node labels to be applied to nodes
+# Format: dictionary of key=value pairs
+# Example:
+# node_labels:
+#   "topology.kubernetes.io/zone": "us-east-1a"
+#   "node.kubernetes.io/instance-type": "m5.large"
+node_labels: {}

roles/kubernetes/node-label/tasks/main.yml

@@ -30,9 +30,8 @@
 - name: Populate inventory node label
   set_fact:
     inventory_node_labels: "{{ inventory_node_labels + ['%s=%s' | format(item.key, item.value)] }}"
-  loop: "{{ node_labels | d({}) | dict2items }}"
+  loop: "{{ node_labels | dict2items }}"
   when:
-    - node_labels is defined
     - node_labels is mapping
 
 - debug:  # noqa name[missing]