Skip to main content

Dagster Tutorial: Building Your First ETL Pipeline

Welcome to this hands-on tutorial where you'll learn how to build an ETL pipeline with Dagster while exploring key parts of Dagster. If you haven't already, please complete the Quick Start tutorial to get familiar with Dagster.

What You'll Learn​

  • Setting up a Dagster project with the recommended project structure
  • Creating Assets and using Resources to connect to external systems
  • Adding metadata to your assets
  • Building dependencies between assets
  • Running a pipeline by materializing assets
  • Adding schedules, sensors, and partitions to your assets

Prerequisites​

Step 1: Set Up Your Dagster Environment​

First, set up a new Dagster project.

  1. Open your terminal and create a new directory for your project:

    Create a new directory
    mkdir dagster-etl-tutorial
    cd dagster-etl-tutorial

  2. Create a virtual environment and activate it:

    Create a virtual environment
    python -m venv venv
    source venv/bin/activate  
    # On Windows, use `venv\Scripts\activate`

  3. Install Dagster and the required dependencies:

    Install Dagster and dependencies
    pip install dagster dagster-webserver pandas

What You've Learned​

Congratulations! You've just built and run your first ETL pipeline with Dagster. You've learned how to:

  • Set up a Dagster project
  • Define Software-Defined Assets for each step of your ETL process
  • Use Dagster's UI to run and monitor your pipeline

Next Steps​

To expand on this tutorial, you could:

  • Add more complex transformations
  • Implement error handling and retries
  • Create a schedule to run your pipeline periodically