Public Docker Images

A collection of custom Docker images published to GitHub Container Registry (GHCR) for personal and community use. These images are automatically built, tested, and published via GitHub Actions workflows.

📦 Available Images

All images are published to ghcr.io/mike12806/public-docker/[image-name] and are available in the following variants:

• latest - Latest build from the main branch
• v*.*.* - Semantic versioned releases
• <git-sha> - Specific commit builds

🏃 Runner

Image: ghcr.io/mike12806/public-docker/runner

A feature-rich self-hosted GitHub Actions runner image designed for Kubernetes environments. This image extends the official GitHub Actions runner with pre-installed tools and dependencies commonly needed in CI/CD workflows.

Base Image: ghcr.io/actions/actions-runner

Included Tools:

• CodeQL - Pre-installed in toolcache layout for security scanning
• AWS CLI v2 - For AWS resource management
• Docker Compose Plugin - For multi-container applications
• Python 3 with pip, venv, and pipx
• Ansible (via pipx) - For configuration management
• Linode CLI (via pipx) - For Linode cloud management
• Backblaze B2 CLI - For B2 cloud storage operations
• yq - YAML processor for parsing and manipulating YAML files
• jq - JSON processor (from base image)
• psql - PostgreSQL client for database operations
• sshpass - Non-interactive SSH password authentication
• rsync - File synchronization utility
• curl, wget - HTTP clients for downloading files
• OpenSSH client - For SSH operations

Key Features:

• CodeQL pre-installed in the standard toolcache layout (/opt/hostedtoolcache/CodeQL)
• Toolcache directory owned by runner user for proper permissions
• Python packages installed via pipx for isolation
• Optimized for self-hosted runners in Kubernetes

Usage Example:

# In your Kubernetes runner deployment
spec:
  containers:
  - name: runner
    image: ghcr.io/mike12806/public-docker/runner:latest

Use Cases:

• Self-hosted GitHub Actions runners
• CI/CD pipelines requiring AWS, Ansible, or CodeQL
• Security scanning workflows with CodeQL
• Infrastructure automation tasks

---

⎈ Kubectl

Image: ghcr.io/mike12806/public-docker/kubectl

A kubectl image with Redis tools for Kubernetes operations that require Redis interactions.

Base Image: docker.io/alpine/kubectl

Included Tools:

• kubectl - Kubernetes command-line tool
• redis-cli - Redis command-line interface for Redis operations
• redis-server - Redis server (available but not typically run in this image)
• curl - HTTP client for API calls and downloads (included in base image)

Usage Example:

# In a Kubernetes Job
apiVersion: batch/v1
kind: Job
metadata:
  name: kubectl-redis-job
spec:
  template:
    spec:
      containers:
      - name: kubectl
        image: ghcr.io/mike12806/public-docker/kubectl:latest
        command: ["kubectl", "get", "pods"]
      restartPolicy: Never

Use Cases:

• Kubernetes maintenance jobs with Redis operations
• CI/CD pipelines for Kubernetes deployments
• Scripted Kubernetes operations requiring Redis client access
• Database migration or initialization scripts that interact with both Kubernetes and Redis

---

🌐 Curl

Image: ghcr.io/mike12806/public-docker/curl

A lightweight Alpine-based image with curl, jq, and bash for making HTTP requests and processing JSON responses.

Base Image: docker.io/alpine

Included Tools:

• curl - Command-line tool for transferring data with URLs
• jq - Lightweight and flexible command-line JSON processor
• bash - Bourne Again SHell for scripting

Usage Example:

# In a Kubernetes Job or CI/CD pipeline
apiVersion: batch/v1
kind: Job
metadata:
  name: api-check
spec:
  template:
    spec:
      containers:
      - name: curl
        image: ghcr.io/mike12806/public-docker/curl:latest
        command: ["sh", "-c"]
        args:
          - |
            curl -s https://api.example.com/data | jq '.results'
      restartPolicy: Never

Use Cases:

• API health checks and monitoring
• JSON data processing in CI/CD pipelines
• Lightweight HTTP client for scripts and automation
• Quick data fetching and transformation tasks

---

🏠 Home Assistant

Image: ghcr.io/mike12806/public-docker/home-assistant

A custom Home Assistant image with DNS fixes for common networking issues. This image patches the aiodns library to disable EDNS cookies by default and optionally disables IPv6 connections to prevent Network unreachable errors on IPv4-only networks.

Base Image: ghcr.io/home-assistant/home-assistant

Key Features:

• EDNS cookie fix — Patches aiodns to disable EDNS cookies by default, preventing DNS timeouts with forwarding resolvers (dnsmasq, pihole, unbound)
• IPv6 disable support — Set DISABLE_IPV6=true to force all connections to IPv4, preventing Network unreachable errors on networks without IPv6 connectivity
• Based on upstream Home Assistant core with minimal modifications
• Compatible with standard Home Assistant configuration and add-ons

Environment Variables:

Variable	Default	Description
DISABLE_IPV6	(unset)	Set to true, 1, or yes to force all network connections to IPv4 only. Prevents [Errno 101] Network unreachable errors when DNS returns AAAA records but the network has no IPv6 connectivity.

Technical Details:

The image applies runtime patches to homeassistant/runner.py at startup:

1. EDNS fix — Wraps the aiodns query method to set EDNS flags to 0 by default. This resolves issues introduced in c-ares 1.33.0+ where EDNS cookies are enabled by default and cause timeouts with certain DNS server configurations.

2. IPv6 disable (opt-in via DISABLE_IPV6) — Monkey-patches socket.getaddrinfo to force AF_INET for all lookups (covers requests, urllib3, and all stdlib-based resolution) and patches aiohttp.TCPConnector to default to AF_INET (covers aiohttp, aiodns, and c-ares async resolution). This is necessary because kernel-level IPv6 disable (sysctl net.ipv6.conf.all.disable_ipv6) is often unavailable in containers, but Python still resolves and attempts AAAA records.

Usage Example:

# In docker-compose.yml or Kubernetes
version: '3'
services:
  homeassistant:
    image: ghcr.io/mike12806/public-docker/home-assistant:latest
    container_name: homeassistant
    privileged: true
    restart: unless-stopped
    environment:
      - TZ=America/New_York
      - DISABLE_IPV6=true  # Optional: force IPv4-only connections
    volumes:
      - ./config:/config
    network_mode: host

# In Kubernetes
env:
- name: TZ
  value: America/New_York
- name: DISABLE_IPV6
  value: "true"

Use Cases:

• Running Home Assistant with DNS servers that don't properly handle EDNS cookies
• Environments using dnsmasq, pihole, or unbound as DNS forwarders
• Networks without IPv6 connectivity (common in home lab Kubernetes clusters)
• Resolving Network unreachable errors for integrations like Tuya, Rheem/EcoNet, Ring (PubNub), HACS, and mobile_app notifications
• Alternative to manual DNS configuration workarounds

---

🔨 Building Images

Images are automatically built and published when:

• Changes are pushed to the Dockerfiles/** directory on the main branch
• Manually triggered via workflow_dispatch
• Daily via scheduled cron job at 10:00 AM UTC

The build process:

1. Discovers all Dockerfiles in the Dockerfiles/ directory
2. Builds each image using Docker Buildx
3. Tags with latest, semantic version, and git SHA
4. Pushes to GitHub Container Registry
5. Creates a GitHub release with changelog

🧪 Testing

Pull requests automatically trigger Docker build verification:

• All Dockerfiles are discovered and built
• Smoke tests are run to verify basic functionality
• Runner image tests verify Python and CodeQL installations
• Other images test basic shell functionality

🔄 Dependency Management

This repository uses Renovate for automated dependency updates:

• Base images in Dockerfiles are automatically updated
• Minor and patch updates are auto-merged
• Major updates require manual review
• Daily checks for new versions

📝 Contributing

To add a new Docker image:

1. Create a new directory under Dockerfiles/[YourImageName]/
2. Add your Dockerfile in that directory
3. Optionally add a README file with image-specific documentation
4. Submit a pull request
5. The CI will automatically build and test your image
6. Once merged, the image will be published to GHCR

📄 License

This repository is maintained by @mike12806. Individual Docker images may contain software with their own licenses. Please review the Dockerfiles for specific licensing information.

🔗 Related Links

• GitHub Container Registry
• GitHub Actions Documentation
• Kubernetes Documentation
• CodeQL Documentation