plastic-yellow-bird 0.2.dev0

No project description provided

Plastic Yellow Bird (aka "PyB"): a Test and Observability Driven Development Framework for AWS CDK

plastic-yellow-bird is a test framework that introduces a whole new approach to monitoring, observability, and infrastructure development. It lets you to accomplish one thing as simply as possible: use the same code you write to verify your infrastructure as your monitoring and alerting stack in production.

By using plastic-yellow-bird, your team can:

• write end-to-end functional and infrastructure tests as easily as you write unit or integration tests
• create AWS Synthetics Canaries with as little boilerplate as possible
• develop a planned production monitoring scheme from the project's outset
• harvest the effort spent writing automated functional tests as zero-cost, thoughtful baseline production monitoring

import plastic_yellow_bird as pyb
import requests

@pyb.canary
def test_api_endpoint_returns_200_with_get(cdk_app):
    response = requests.get(cdk_app.api_stack.api.url)
    assert response.status_code == 200

Imagine you create the above test file in a CDK project directory that contains an Amazon API Gateway CDK app. By using plastic-yellow-bird, which is based on pytest, simply running pytest, plastic-yellow-bird would deploy your CDK app and launch the test_api_endpoint_returns_200_with_get code as an AWS Synthetics Canary in the same account, then wait for the Canary to run and report a status. plastic-yellow-bird would then report this test's Pass or Fail state just like any other pytest test result, based on the status of the Canary. In other words, if the Canary launched by plastic-yellow-bird succeeded, your test would be reported as passing, and if the Canary failed, the test would be reported as failing.

Design Principles

• Most use cases should require nothing more than test code and a decorator to indicate a PyB test
  • "Make simple things simple, and complex things possible" - Alan Kay
• Code should be beautiful
• Developers are smart and capable of doing incredibly difficult things, but using this tool shouldn't be one of those incredibly difficult things
• If no one uses this in real projects, we've failed