Skip to content

Latest commit

 

History

History
68 lines (42 loc) · 3.49 KB

airflow-tutorial.md

File metadata and controls

68 lines (42 loc) · 3.49 KB
Raw

Canarypy Airflow Tutorial

This page provides a tutorial on how to use canarypy with Apache Airflow. It assumes that you have a working knowledge of Airflow.

Environment setup

  1. From the root of the examples directory, run docker-compose up -d to start the canarypy backend. This will create the folowing containers:
  • API: The canarypy backend API.
  • DB: The canarypy backend database.
  • WEB: The canarypy backend dashboard.
  • INIT: Temporary container that runs the canarypy backend database migrations.
  1. From the root of the examples/airflow directory, run docker-compose up -d to start the Airflow instance. This will create the local Airflow setup with the canarypy plugin package already installed.

If you can access http://localhost:8080, then the Airflow instance is running correctly. The airflow user/passwrod is: airflow/airflow.

In the Airflow DAG list, you should see a DAG called docker_operator_demo. This DAG contains three DockerOperator tasks that uses the python image to print basic commands. On this tutorial, we'll create a few releases to demo the version switching automatically.

Canarypy setup

The very first thing to do within Canarypy is to create a Product. In our example is Python.

1. Create a product in the canarypy backend

The canarypy command-line interface (CLI) provides commands to interact with the canarypy backend. The product create command is used to create a new product.

Before running the command, we need to specify the canarypy backend URL by setting the CANARYPY_URL environment variable. This can be done directly in the command line like so:

export CANARYPY_URL=http://localhost:9090

This sets the backend URL to http://localhost:9090. Ensure that your canarypy backend is running on this URL. If it's running on a different URL, replace http://localhost:9090 with the correct URL.

After setting the CANARYPY_URL, you can create the product:

canarypy product create

This command will prompt you a few questions:

  • Enter the name of this product? super python
  • What the git repository url of this product? https://github.com/python/cpython
  • What the artifact url of this product? python

The artifact url is what it will be used by Canarypy Airflow plugin to fetch the correct artifact.

2. Create a release for the product

The canarypy release create command is used to create a new release for a product. This command takes a --semver-version argument to specify the semantic version of the release, and an --artifact-url argument to specify the URL where the release's artifact can be downloaded.

In a real world scenario, the canarypy release create command can be integrated into your CI/CD pipeline. After your product is built and deployed, the command can be executed to create a new release in the canarypy backend:

canarypy release create --semver-version 3.10-bullseye --artifact-url python

After creating the release, trigger the DAG in the Airflow UI at http://localhost:8080. You can check the logs of the Docker operator to confirm that it's using the correct version.

5. Create another release and check the DAG running on the newest version

To create another release for your product, use the canarypy release create command again with the new version and artifact URL:

canarypy release create --semver-version 3.11-bullseye --artifact-url python

Now check that the DAG will run using the newest release version of your product.