Skip to content

cfa-cloudops Overview

Outline

Description

The cfa-cloudops python module is intended to ease the challenge of working in the cloud (currently limited to Azure) for data scientists at CFA. Typically, it takes deep knowledge to authenticate and complex coding patterns to interact with the cloud via python, which takes away time and resources from data scientists doing predictions, modeling, and more impactful work. cfa-cloudops simplifies many repeated workflows when interacting with the cloud and unifies CFA's approach in developer interaction with the cloud. For example, creating a pool and running jobs in Azure may take several credentials and different clients to complete, but with cfa-cloudops, this is reduced to a small number of functions with only a few user-provided parameters.

Getting Started

To use cfa-cloudops, you need Python 3.9 or higher, Azure CLI, and any python package manager.

Currently cfa-cloudops is only available to install from GitHub.

To install using pip: 

pip install git+https://github.com/CDCgov/cfa-cloudops.git

Components of cfa-cloudops

There are several components of this repo that provide benefits to developers interacting with the Cloud (currently Azure only). They are:

  • low-level python functions to assist in cloud interaction
    • auth
    • blob
    • client
    • endpoints
    • job
    • task
  • CloudClient object for easy interaction with the cloud
    • more info found here
  • automation component to run jobs/tasks from a configuration file