Configuring Airflow

The chart allows for setting arbitrary Airflow configuration in values under the config key. Some of the defaults in the chart differ from those of core Airflow and can be found in values.yaml.

As an example of setting arbitrary configuration, the following yaml demonstrates how one would allow webserver users to view the config from within the UI:

config:
  webserver:
    expose_config: 'True'  # by default this is 'False'

Generally speaking, it is useful to familiarize oneself with the Airflow configuration prior to installing and deploying the service.

Note

The recommended way to load example DAGs using the official Docker image and chart is to configure the AIRFLOW__CORE__LOAD_EXAMPLES environment variable in extraEnv (see Parameters reference). The official Docker image has AIRFLOW__CORE__LOAD_EXAMPLES=False set within the image, so you need to override it with an environment variable when deploying the chart in order for the examples to be present.

Note

The AIRFLOW__DATABASE__LOAD_DEFAULT_CONNECTIONS variable is not used by the Chart. Airflow Helm Chart is intended to be used as production deployment and loading default connections is not supposed to be handled during Chart installation. The Chart is intended to install and configure the Apache Airflow software and create database structure, but not to fill-in the data which should be managed by the users. The default connections are only meaningful when you want to have a quick start with Airflow or do some development and adding the data via Helm Chart installation is not a good idea.

Was this entry helpful?