Skip to content

About

Update and Deploy GPU-CI AMI on AWS

Resources

Code of conduct

Security policy

Stars

1 star

Watchers

1 watching

Forks

Latest commit

 

History

17 Commits

Folders and files

Repository files navigation

GPU CI Configuration

This repository manages GPU-enabled AMI images for CI/CD workflows with cirun.io using Packer and GitHub Actions.

Setup Requirements

AWS Configuration

  1. Configure AWS IAM role with permissions for:

    • EC2 instance management
    • AMI creation/deletion
    • Image sharing across regions
  2. Set up GitHub repository secrets:

    • AWS_ROLE_ARN - IAM role ARN for AWS authentication
    • BOT_GH_TOKEN - GitHub Personal Access Token with repo permissions

Repository Access

The workflow updates AMI configurations in these repositories:

  • scverse/anndata

Ensure the BOT_GH_TOKEN token has write access to these repositories.

Usage

Build New AMI

Run the "Build AMI" workflow to create a new GPU-enabled AMI with:

  • Ubuntu 24.04 LTS
  • Docker & Docker Compose
  • NVIDIA drivers
  • Pre-installed development tools

Update CI Configurations

Run the "Update CI AMI" workflow to:

  • Automatically detect the latest AMI
  • Create PRs in target repositories to update .cirun.yml files
  • Update AMI references across multiple projects

AMI Specifications

  • Base OS: Ubuntu 24.04 LTS
  • Instance Type: GPU-enabled (g4dn.xlarge for building)
  • Storage: 125GB gp3 with 6,000 IOPS
  • Tools: Docker, NVIDIA drivers, development utilities

About

Update and Deploy GPU-CI AMI on AWS

Resources

Code of conduct

Security policy

Stars

1 star

Watchers

1 watching

Forks

Releases

Packages

Used by

Contributors

Languages