This component is responsible to write data into the ACONA Data Warehouse. It schedules jobs and provides functionality to calculate scores, recommendations, forecasts and alerts.
Scheduling is done with Apache Airflow: https://airflow.apache.org/
For forecasts Prophet is used, an Open Source software by Facebook: https://facebook.github.io/prophet/
ACONA stands for Augmented Content Analytics - an open source tool that automatically analyzes and simplifies data, for example from server logs or existing (open source) analytics tools, and proposes concrete measures for optimizing content.
More in the general documentation: https://app.gitbook.com/@acolono/s/acona/
Build the docker image
docker build . -f Dockerfile --tag acona-airflow:0.0.1
Rename .env_template to .env, create a fernet key and update values in .env.
How to generate a fernet key: https://airflow.apache.org/docs/apache-airflow/stable/security/secrets/fernet.html
Run docker-compose
docker-compose up
Before you can run a dag you need to create the following variables:
- WAREHOUSE_URL: The ACONA Warehouse URL you want to use
- WAREHOUSE_TOKEN: The authentication token for the ACONA Warehouse and
- MATOMO_URL and MATOMO_TOKEN for every configured domain like this: When your Domain ID is 1 name the variables MATOMO_URL_1 and MATOMO_TOKEN_1
You also need to configure an airflow postgres connection to your ACONA Data Warehouse, named "acona_data_warehouse".
To set up your own ACONA Data Warehouse, see https://github.com/acolono/ACONA-data-warehouse