Packaging¶

Overview¶

Geniusrise uses docker for packaging and delivering modules.

Capabilities¶

Docker Image Creation: Create Docker images with custom base images, working directories, and local directories.
Package Installation: Install both OS-level and Python packages during the Docker image creation.
Environment Variables: Set environment variables in the Docker container.
Multi-Repository Support: Upload Docker images to multiple types of container repositories.
Authentication: Supports various authentication methods for different container repositories.

Command-Line Interface¶

Syntax¶

genius docker package <image_name> <repository> [options]

Parameters¶

<image_name>: The name of the Docker image to build and upload.
<repository>: The container repository to upload to (e.g., "ECR", "DockerHub", "Quay", "ACR", "GCR").

Options¶

--auth: Authentication credentials as a JSON string. Default is an empty JSON object.
--base_image: The base image to use for the Docker container. Default is "nvidia/cuda:12.2.0-runtime-ubuntu20.04".
--workdir: The working directory in the Docker container. Default is "/app".
--local_dir: The local directory to copy into the Docker container. Default is ".".
--packages: List of Python packages to install in the Docker container. Default is an empty list.
--os_packages: List of OS packages to install in the Docker container. Default is an empty list.
--env_vars: Environment variables to set in the Docker container. Default is an empty dictionary.

Authentication Details¶

ECR: {"aws_region": "ap-south-1", "aws_secret_access_key": "aws_key", "aws_access_key_id": "aws_secret"}
DockerHub: {"dockerhub_username": "username", "dockerhub_password": "password"}
ACR: {"acr_username": "username", "acr_password": "password", "acr_login_server": "login_server"}
GCR: {"gcr_key_file_path": "/path/to/keyfile.json", "gcr_repository": "repository"}
Quay: {"quay_username": "username", "quay_password": "password"}

Examples¶

Uploading to ECR (Amazon Elastic Container Registry)¶

genius docker package geniusrise ecr --auth '{"aws_region": "ap-south-1"}'

Uploading to DockerHub¶

genius docker package geniusrise dockerhub --auth '{"dockerhub_username": "username", "dockerhub_password": "password"}'

Uploading to ACR (Azure Container Registry)¶

genius docker package geniusrise acr --auth '{"acr_username": "username", "acr_password": "password", "acr_login_server": "login_server"}'

Uploading to GCR (Google Container Registry)¶

genius docker package geniusrise gcr --auth '{"gcr_key_file_path": "/path/to/keyfile.json", "gcr_repository": "repository"}'

Uploading to Quay¶

genius docker package geniusrise quay --auth '{"quay_username": "username", "quay_password": "password"}'

Uploading with Custom Packages and OS Packages¶

genius docker package geniusrise dockerhub \
    --packages geniusrise-listeners geniusrise-databases geniusrise-huggingface geniusrise-openai \
    --os_packages libmysqlclient-dev libldap2-dev libsasl2-dev libssl-dev

Uploading with Environment Variables¶

genius docker package geniusrise dockerhub --env_vars '{"API_KEY": "123456", "ENV": "production"}'

Complex Examples¶

1. Uploading to ECR with Custom Base Image and Packages¶

This example demonstrates how to upload a Docker image to ECR with a custom base image and additional Python packages.

genius docker package my_custom_image ecr \
    --auth '{"aws_region": "us-west-2", "aws_secret_access_key": "aws_key", "aws_access_key_id": "aws_secret"}' \
    --base_image "python:3.9-slim" \
    --packages "numpy pandas scikit-learn" \
    --os_packages "gcc g++"

2. Uploading to DockerHub with Environment Variables and Working Directory¶

This example shows how to upload a Docker image to DockerHub with custom environment variables and a specific working directory.

genius docker package my_app dockerhub \
    --auth '{"dockerhub_username": "username", "dockerhub_password": "password"}' \
    --env_vars '{"DEBUG": "True", "SECRET_KEY": "mysecret"}' \
    --workdir "/my_app"

3. Uploading to ACR with Multiple Local Directories¶

In this example, we upload a Docker image to Azure Container Registry (ACR) and specify multiple local directories to be copied into the Docker container.

# First, create a Dockerfile that copies multiple directories
# Then use the following command
genius docker package multi_dir_app acr \
    --auth '{"acr_username": "username", "acr_password": "password", "acr_login_server": "login_server"}' \
    --local_dir "./app ./config"

4. Uploading to GCR with Custom Base Image, Packages, and OS Packages¶

This example demonstrates how to upload a Docker image to Google Container Registry (GCR) with a custom base image, Python packages, and OS packages.

genius docker package my_ml_model gcr \
    --auth '{"gcr_key_file_path": "/path/to/keyfile.json", "gcr_repository": "repository"}' \
    --base_image "tensorflow/tensorflow:latest-gpu" \
    --packages "scipy keras" \
    --os_packages "libsm6 libxext6 libxrender-dev"

5. Uploading to Quay with All Customizations¶

This example shows how to upload a Docker image to Quay with all available customizations like base image, working directory, local directory, Python packages, OS packages, and environment variables.

genius docker package full_custom quay \
    --auth '{"quay_username": "username", "quay_password": "password"}' \
    --base_image "alpine:latest" \
    --workdir "/custom_app" \
    --local_dir "./src" \
    --packages "flask gunicorn" \
    --os_packages "bash curl" \
    --env_vars '{"FLASK_ENV": "production", "PORT": "8000"}'