chiltepin 0.1.4

Federated NWP Workflow Tools

Chiltepin

Overview

This repository is a collection of tools and demonstrations used to explore and test various technologies for implementing exascale scientific workflows. This collection of resources is not intended for production use, and is for research purposes only.

Chiltepin provides Python decorators and utilities for building scientific workflows that can execute on distributed computing resources using Parsl and Globus services.

Platform Support

Chiltepin is developed and tested on Linux:

• ✅ Linux: Full support for all features
• 🍎 macOS: Supported for task submission and data transfer (endpoint management not available)
• ❌ Windows: Native execution not supported; Chiltepin's use of Parsl relies on fork-based multiprocessing
• 🐳 Docker: Full Linux-based test suite and feature support available via container on all platforms

Documentation

📚 Full documentation is available at Read the Docs

Key documentation sections:

• Installation Guide - Installing Chiltepin
• Quick Start - Your first Chiltepin workflow
• Tasks - Python, Bash, and Join task decorators
• Configuration - Configuring compute resources
• Endpoints - Managing Globus Compute endpoints
• Data Transfer - Using Globus for data movement
• Testing Guide - Running the test suite

Quick Start

Install Chiltepin using pip:

pip install chiltepin

For detailed installation instructions including conda, Docker, and platform-specific guidance, see the Installation Guide.

Contributing

Contributions are welcome! For development installation and running tests, clone the repository and install in editable mode:

git clone https://github.com/NOAA-GSL/chiltepin.git
cd chiltepin
python -m venv .chiltepin
source .chiltepin/bin/activate
pip install -e ".[test]"

See the Testing Guide for more information.

License

See LICENSE for details.