Pipeline Abstraction is built on task group, it is designed to build abstractions on a group of re-usable tasks and their structure.
This repo could be separated into two parts:
- The Airflow Infra described below
- The demo Airflow dags
- Install Pants: https://www.pantsbuild.org/docs/installation
- Install Python 3.10.x
- Because in pants.toml, we set the intepreter to Python 3.10.x
- Because on Ubuntu 22.04, the default Python is Python 3.10.x
- No Python Virtual Environment needed
Here is a recommended way to install Python 3.10.x:
bin/install_python
bin/install_airflow && bin/airflow_config && airflow standalone
bin/install_airflow
bin/airflow_config
It will:
- Set
core.dags_folder
to the PEX user code dynamically - Set
core.load_examples
toFalse
airflow standalone
If the command line airflow
is not available, please add $HOME/bin
to $PATH
.
For more info, see https://airflow.apache.org/docs/apache-airflow/stable/start.html
Juse remove the $AIRFLOW_HOME:
rm -rf $HOME/airflow