Cryo Base Image

This repository contains the base Docker image with pre-compiled Cryo binary for fast application builds.

Purpose

Building Cryo from source takes several hours, especially for ARM64 architecture. This base image is built weekly (or on-demand) and contains:

• Pre-compiled Cryo binary (specific commit: eba6192298e40add4d35d1587511d875f6d770e4)
• Custom patches for enhanced functionality
• Python 3 runtime
• All necessary system dependencies (libssl, gcc, g++)

Usage

In your application's Dockerfile:

FROM ghcr.io/gnosischain/cryo-base:latest

# Your application code here
WORKDIR /app
COPY . .
# ...

One can run this version as

docker run --rm -v "$(pwd)":/data cryo-base:latest \
  cryo blocks \
  --blocks 39828831:39828831 \
  --rpc 'URL' \
  --include-columns block_number timestamp withdrawals_root withdrawals \
  -o /data

Applied Patches

This image includes a modular patching system. Current patches:

Patch	Description
01_null_traces.py	Makes trace functions null-safe for RPCs that return null instead of []
02_withdrawals.py	Adds withdrawals column support to blocks dataset

Adding a New Patch

1. Create a new file in patches/ with the naming pattern NN_patch_name.py (numeric prefix controls execution order)

2. Use this template structure:

   #!/usr/bin/env python3
   from pathlib import Path
   
   DESCRIPTION = "Short description of what this patch does"
   TARGET_FILE = Path("crates/freeze/src/path/to/file.rs")
   
   def main():
       # Return True (success), False (failure), or "skipped" (already applied)
       ...

3. The patch runner (apply_patches.py) will automatically discover and apply patches in alphabetical order

4. Build the image to apply:

   docker build -t cryo-base .

See patches/_template.py.example for a full template.

Directory Structure

.
├── Dockerfile
├── README.md
├── apply_patches.py      # Patch runner script
└── patches/
    ├── 01_null_traces.py
    ├── 02_withdrawals.py
    └── _template.py.example

Build Schedule

• Automatic builds: Every Sunday at 2 AM UTC
• Manual builds: On push to main branch or via workflow dispatch

Manual Build

To trigger a manual build with a specific Cryo commit:

1. Go to Actions → "Build Cryo Base Image"
2. Click "Run workflow"
3. Optionally specify a Cryo commit hash
4. Click "Run workflow"

Versioning

• latest: Always points to the most recent build
• YYYYMMDD: Date-based tags for specific versions

Architecture Support

• linux/amd64
• linux/arm64

Both architectures are built in parallel for faster builds.

Update Process

When Cryo releases a new version:

1. Update the commit hash in Dockerfile
2. Push to main branch or trigger manual build
3. Wait for build completion (usually 2-3 hours)
4. Update your application to use the new base image

To add a new patch:

1. Add your patch file to patches/ directory
2. Follow the naming convention NN_description.py
3. Implement DESCRIPTION, TARGET_FILE, and main() function
4. Push to main branch or trigger manual build

Contents

The base image includes:

• Debian bullseye-slim base
• Cryo binary at /usr/local/bin/cryo (with custom patches)
• Python 3.x with pip
• Build essentials (gcc, g++)
• OpenSSL libraries
• CA certificates