Image Comparison Matrix¶

Complete comparison of all image variants with detailed tool versions, sizes, and use case recommendations.

Quick Comparison¶

Complete Tool Matrix¶

Infrastructure as Code¶

All variants include: Full Terraform ecosystem for infrastructure as code.

Kubernetes Tools¶

All variants include: Complete Kubernetes toolchain for cluster management.

Configuration Management¶

All variants include: Ansible for agentless configuration management.

Security & Scanning¶

All variants include: Trivy for container and IaC security scanning.

AWS-Specific Tools¶

Available in: aws-devops, all-devops only

GCP-Specific Tools¶

Available in: gcp-devops, all-devops only

Development Tools¶

All variants include: Python 3.12 and Node.js 20 LTS for scripting and automation.

AI CLI Tools¶

All variants include: Complete AI CLI toolkit for AI-assisted DevOps workflows.

Version Control & CI/CD¶

All variants include: Git and GitHub CLI for version control workflows.

Utilities¶

All variants include: Essential command-line utilities.

Size Breakdown by Category¶

Base Image (~2.0 GB)¶

pie title Base Image Components
    "Rocky Linux Base" : 500
    "IaC Tools" : 125
    "Kubernetes Tools" : 110
    "Python + Node.js" : 240
    "Security (Trivy)" : 240
    "Ansible" : 95
    "System Packages" : 300
    "Utilities" : 390

AWS Layer (+800 MB)¶

GCP Layer (+900 MB)¶

All-DevOps (+1.2 GB)¶

Not simply AWS + GCP (would be 1.7 GB)

Optimizations: - Shared Python runtime: Save ~50 MB - Shared dependencies: Save ~100 MB - Single-layer cleanup: Save ~150 MB

Actual addition: 1.2 GB (save 500 MB through optimization)

Performance Comparison¶

Build Times¶

Notes: - Warm builds benefit from cached base layer - Multi-arch builds run sequentially (amd64, then arm64) - Can parallelize multi-arch builds with multiple runners

Pull Times (1 Gbps network)¶

Factors affecting pull time: - Network speed - Registry location (GHCR generally faster from GitHub Actions) - Layer caching (subsequent pulls are faster)

Startup Times¶

Note: Startup times measured on M2 MacBook Pro (arm64). Times may vary by host.

Use Case Recommendations¶

Platform-Only Workflows¶

Typical usage:

# Base is a local Dockerfile target (not a published registry image)
docker build --target base -t devops-base:local .
docker run -it --rm devops-base:local

AWS-Centric Workflows¶

Typical usage:

docker pull ghcr.io/jinalshah/devops/images/aws-devops:latest

Benefits over all-devops: - 400 MB smaller (13% reduction) - Faster pulls and builds - No unused GCP tools

GCP-Centric Workflows¶

Typical usage:

docker pull ghcr.io/jinalshah/devops/images/gcp-devops:latest

Benefits over all-devops: - 300 MB smaller (9% reduction) - Faster pulls and builds - No unused AWS tools

Multi-Cloud Workflows¶

Typical usage:

docker pull ghcr.io/jinalshah/devops/images/all-devops:latest

Trade-offs: - Largest image (3.2 GB) - Complete toolset for both clouds - Unified image across teams

Architecture Support¶

Multi-Architecture Matrix¶

Docker automatically selects the correct architecture based on your platform.

Manifest list:

docker manifest inspect ghcr.io/jinalshah/devops/images/all-devops:latest

Shows available architectures for automatic selection.

Registry Availability¶

Image Availability by Registry¶

Recommended: Use GHCR (GitHub Container Registry) for: - No rate limits on public images - Better performance from GitHub Actions - Built-in security scanning

Decision Tree¶

graph TD
    A[Which image?] --> B{Need cloud<br/>CLIs?}
    B -->|No| C[Base Image<br/>~2.0 GB]
    B -->|Yes| D{Which<br/>cloud?}
    D -->|AWS only| E[aws-devops<br/>~2.8 GB]
    D -->|GCP only| F[gcp-devops<br/>~2.9 GB]
    D -->|Both| G[all-devops<br/>~3.2 GB]
    D -->|Unsure| G

    style C fill:#90EE90
    style E fill:#FFD700
    style F fill:#FFD700
    style G fill:#FFA500

Cost Considerations¶

Storage Costs¶

Assuming $0.10/GB/month for container registry storage:

For 10 developers pulling daily: - Bandwidth: ~10 GB/day per image - Monthly bandwidth: ~300 GB - Bandwidth cost: ~$30/month (varies by registry)

Optimization tip: Use smaller images (aws-devops or gcp-devops) to reduce bandwidth costs.

CI/CD Pipeline Costs¶

GitHub Actions (2,000 free minutes/month):

Savings: Using aws-devops vs all-devops saves ~6 minutes/month per 100 builds.

Migration Guide¶

From Docker Hub to GHCR¶

Old:

docker pull js01/all-devops:latest

New:

docker pull ghcr.io/jinalshah/devops/images/all-devops:latest

CI/CD update:

# .github/workflows/ci.yml
container:
  image: ghcr.io/jinalshah/devops/images/all-devops:1.0.abc1234

From all-devops to Cloud-Specific¶

Step 1: Determine cloud usage

# Check if AWS CLI is used
grep -r "aws " .github/

# Check if gcloud is used
grep -r "gcloud " .github/

Step 2: Update image

# If only AWS found
container:
  image: ghcr.io/jinalshah/devops/images/aws-devops:1.0.abc1234

# If only GCP found
container:
  image: ghcr.io/jinalshah/devops/images/gcp-devops:1.0.abc1234

Step 3: Test and measure - Verify all workflows pass - Measure pull time improvement - Monitor build duration

Next Steps¶

• Base Layer Details - Tool breakdown
• Cloud Layer Details - AWS and GCP additions
• Building Images - Customize your own
• Use Images Guide - Get started