# OKD Cluster Automation

This repository contains Ansible automation for deploying and managing OKD clusters. The automation supports both initial cluster provisioning and day-2 operations.

## Directory Structure

```
ansible/
├── inventories/
│   └── hub/
│       └── group_vars/
│           └── okd_cluster.yml    # Cluster-specific variables
├── playbooks/
│   ├── provision_cluster.yml      # Initial cluster deployment
│   └── day2_operations.yml        # Post-deployment configurations
└── roles/
    └── okd_install/              # OKD installation role
```

## Prerequisites

- Ansible 2.9 or higher
- Python 3.6 or higher
- `ansible-navigator` for container-based execution
- Access to target machines with SSH key authentication
- DNS configured for the cluster domain
- Required network configuration (DHCP, load balancers, etc.)

## Configuration

The cluster configuration is defined in `inventories/hub/group_vars/okd_cluster.yml`. Key configurations include:

```yaml
# OKD cluster configuration
okd_version: "4.17.0-okd-scos.0"
base_domain: "okd.container.mom"
okd_name: "hub"
```

## Usage

### Initial Cluster Provisioning

To provision a new cluster:

```bash
# Make sure you're in the ansible directory
cd ansible

# Run the playbook
ansible-playbook playbooks/provision_cluster.yml --inventory inventories/hub
```

This will:
1. Download and install OKD installation tools
2. Generate installation configurations
3. Create the cluster
4. Configure hybrid master nodes (control plane nodes are schedulable)

### Installation Process

The installation process follows these steps:

1. **Prerequisites Check and Setup**
   - Verifies and installs required tools (openshift-install, oc)
   - Creates necessary directories

2. **Configuration Generation**
   - Creates install-config.yaml from template
   - Generates Kubernetes manifests
   - Creates Ignition configs

3. **Cluster Installation**
   - Initiates the cluster creation
   - Waits for bootstrap completion
   - Configures master nodes as schedulable (hybrid mode)
   - Waits for installation completion

## Available Tags

- `prereq`: Install prerequisites
- `config`: Generate configurations
- `install`: Run the installation
- `ignition`: Generate ignition configs

## Node Configuration

The automation supports a hybrid node setup where master nodes are also workers:

```yaml
# Node configuration
okd_controller_replicas: 3  # 3 master nodes
okd_worker_replicas: 0      # No dedicated workers
okd_make_masters_schedulable: true  # Hybrid mode

# Current node configuration
okd_masters:
  - name: master0
    ip: "188.244.117.200"
  - name: master1
    ip: "188.244.117.223"
  - name: master2
    ip: "188.244.117.142"
```

## Important Notes

1. **Role Path**
   - Ensure you run the playbook from the `ansible` directory
   - The roles directory must be at `ansible/roles`
   - Verify role path with: `ansible-config dump | grep ROLES_PATH`

2. **Installation Monitoring**
   - Monitor progress: `tail -f {{ okd_cluster_artifacts }}/.openshift_install.log`
   - Bootstrap node access: `ssh -o UserKnownHostsFile=/dev/null core@{{ okd_bootstrap_external_static_ip }}`

3. **Network Requirements**
   - Ensure VIPs are available
   - DNS must be properly configured
   - Network CIDR must be correctly set

## Troubleshooting

Common issues and solutions:

1. **Role Not Found**
   ```bash
   # Fix by creating ansible.cfg in the ansible directory
   echo "[defaults]
   roles_path = ./roles" > ansible.cfg
   ```

2. **Network Issues**
   - Verify VIP accessibility
   - Check DNS resolution
   - Confirm network CIDR is correct

3. **Node Access**
   - Verify SSH key authentication
   - Check network connectivity
   - Ensure proper permissions

## License

This project is licensed under the MIT License.