v1.3.9

Added

• talos-image module: SBC overlay support for single-board computers
  • New sbc_overlay variable for board-specific overlays (turingrk1, rpi_generic, rock5b, etc.)
  • Auto-detection of overlay images from overlay name
  • Supports 16+ SBC boards across Rockchip, Raspberry Pi, Jetson, and Allwinner families
• talos-cluster module: Added talosconfig_path output
• talos-wipe.sh: Added --yes / -y flag for non-interactive automation

Fixed

• talos-cluster module: Fixed talosconfig output format - now generates proper YAML with context, endpoints, nodes (previously output unusable JSON)
• talos-wipe.sh: Removed eMMC wipe attempt - eMMC is system disk and cannot be wiped via talosctl reset

Changed

• Updated README version references to ~> 1.3.9

Verified

• All addon modules tested on Talos v1.9.2 (Turing RK1 with turingrk1 overlay)
  • metallb: L2 mode with IP pool 10.10.88.80-89
  • ingress-nginx: LoadBalancer service on 10.10.88.80
  • cert-manager: Self-signed CA and ClusterIssuers ready
  • longhorn: StorageClass created with NVMe storage class
  • monitoring: Prometheus + Grafana with persistent storage
  • portainer: Agent accessible on 10.10.88.81:9001
• Talos Image Factory integration with SBC overlays verified

v1.3.8

Changes

• Wipe scripts now wipe both NVMe and eMMC by default - Both drives are wiped to ensure clean state for re-flashing
• Prominent data destruction warning - Clear visual warning showing exactly what will be destroyed
• Safer confirmation - Changed from 'yes' to 'DESTROY' to prevent accidental execution
• Skip flags available - Use --no-nvme or --no-emmc to skip specific drives

Example Output

╔════════════════════════════════════════════╗
║       ⚠️  WARNING: DATA DESTRUCTION ⚠️     ║
╚════════════════════════════════════════════╝

Data to be PERMANENTLY DESTROYED:
  • NVMe drives: /dev/nvme0n1
  • eMMC boot drives: /dev/mmcblk0
  • K3s data directories
  • Kubernetes state and configurations

This will PERMANENTLY DESTROY ALL DATA on these nodes!
Nodes will be UNBOOTABLE until re-flashed via BMC!

Type 'DESTROY' to confirm:

Full Changelog: v1.3.7...v1.3.8

v1.3.7

Bug Fixes

• Helper scripts bash compatibility - Fixed set -e compatibility issues with STEP increment and log_output function that caused scripts to exit prematurely
• Secrets auto-loading - Scripts now auto-load credentials from ~/.secrets/turning-pi-cluster-bmc file format (ip, username, password)
• SSH key detection - Scripts auto-detect SSH key from ~/.secrets/turningpi-cluster

Tested

All scripts verified working on physical TuringPi cluster:

• cluster-preflight.sh - 18 checks passed
• k3s-wipe.sh - Dry-run completed successfully
• talos-wipe.sh - Dry-run completed successfully

Full Changelog: v1.3.6...v1.3.7

v1.3.6

What's New

Added

• talos-image module - Generate Talos images with extensions (iscsi-tools, util-linux-tools) for Longhorn support
• docs/WORKFLOWS.md - Complete cluster lifecycle documentation with Mermaid flowcharts for Talos and K3s
• scripts/cluster-preflight.sh - Pre-deployment validation script checking tools, BMC connectivity, node status
• scripts/talos-wipe.sh - Enhanced Talos cluster wipe with env vars, credential files, terraform cleanup, force power-off
• scripts/k3s-wipe.sh - Enhanced K3s cluster wipe with node draining, container cleanup, iptables cleanup

Changed

• Updated talos-full-stack example to use talos-image module for automatic image generation
• Enhanced README with documentation links and helper script examples
• Added platform-specific configurations to addon modules (Talos vs K3s/Armbian)
• Added storage capacity planning guidance for eMMC-constrained nodes

Full Changelog: v1.3.5...v1.3.6

v1.3.5

What's Changed

Added

• cert-manager addon module - TLS certificate management with Let's Encrypt and self-signed CA support
• docs/UPGRADE.md with comprehensive upgrade guidance
• namespace variable to all addon modules (metallb, ingress-nginx, longhorn, monitoring, portainer)
• controller_resources and speaker_resources to MetalLB module
• controller_replicas, controller_resources, enable_metrics to ingress-nginx module
• manager_resources, ui_replicas to Longhorn module
• replicas variable to Portainer module
• Grafana password validation (minimum 8 characters) in monitoring module

Changed

• All addon modules now use configurable namespaces instead of hardcoded values
• Improved resource configuration flexibility across all addon modules

Fixed

• MetalLB and cert-manager modules now use values block instead of set blocks for Helm provider v3.x compatibility

Full Changelog: v1.3.4...v1.3.5

v1.3.4

What's Changed

Changed

• Synchronized release with terraform-provider-turingpi v1.3.4
• Provider now supports BMC firmware 2.3.4 API response format

Provider Compatibility

This release is synchronized with terraform-provider-turingpi v1.3.4, which includes:

• Full compatibility with BMC firmware 2.3.4 API response format
• All data sources (turingpi_info, turingpi_about, turingpi_usb, turingpi_power, turingpi_sdcard) working correctly

Full Changelog: v1.3.3...v1.3.4

v1.3.2

Changed

• Bump actions/checkout from v4 to v6
• Bump terraform-linters/setup-tflint from v4 to v6

v1.3.1

Added

• README badges (CI status, Terraform Registry, License) to root and all submodule READMEs

v1.3.0

Added

CI/CD & Automation

• GitHub Actions workflow for Terraform validation (fmt, init, validate) on PRs
• TFLint integration with recommended ruleset (.tflint.hcl)
• Trivy security scanning for misconfigurations (trivy.yaml)
• terraform-docs integration for auto-generated documentation (.terraform-docs.yml)
• Dependabot for Terraform provider and GitHub Actions updates
• Pre-commit hooks for local validation (.pre-commit-config.yaml)
• CODEOWNERS file for automatic PR review requests

Repository Configuration

• Branch protection with required status checks and code owner reviews
• Issue templates (bug report, feature request)
• Pull request template with validation checklist
• CONTRIBUTING guide with development setup instructions

Removed

• Unused install_timeout variable from k3s-cluster module
• Unused allow_scheduling_on_control_plane variable from talos-cluster module

v1.2.4

Added

• talos_version variable to talos-cluster module for explicit Talos version in config generation
• kubernetes_version variable to talos-cluster module for explicit Kubernetes version