Deploying Dagster

Getting started with Dagster is as easy as running:

pip install dagit

Dagster supports a wide variety of more sophisticated deployment strategies for production. Typically, you'll want to deploy Dagit somewhere, configure the way that it stores information and launches runs, and make sure your pipelines are configured to run appropriately for your environment.

For example, if you're using the Celery engine, you'll want to run worker processes to parallelize execution. Alternatively, you may want to deploy Dagster pipelines into a completely different scheduling and execution environment, like Apache Airflow.

A good place to start is the Dagster instance page, which showcases (1) which configurations should be set globally for your Dagster deployment and (2) which configurations are set per pipeline.