E-Commerce Data Warehouse ETL

This project implements an ETL (Extract, Transform, Load) process to migrate and transform data from an OLTP (Online Transaction Processing) system to a star schema in a data warehouse. The ETL process is written in Kotlin and Spark, and it is orchestrated using Apache Airflow.

Features

• ETL Processing of MySQL tables:

  • T_CATEGORY: product categories.
  • T_CUSTOMER: customer data and their addresses.
  • T_ORDER: information about orders.
  • T_ORDER_REL: information about products in orders.
  • T_PRODUCT: Prosses product information.
  • T_PROMO and T_PROMO_REL: information about promotions and the products affected by them.

• Data Transformation:

  • Transforms data from OLTP format to a star schema suitable for analytical queries.

• Orchestration:

  • Utilizes Apache Airflow for scheduling and managing the ETL workflows.

Technologies Used

• Programming Language: Kotlin
• Data Processing Framework: Apache Spark v3.3.2
• Workflow Orchestration: Apache Airflow

Getting Started

Prerequisites

• Kotlin: Ensure you have Kotlin installed. Install Kotlin.
• Apache Spark v3.3.2: Ensure you have Apache Spark installed. Install Spark.
• Apache Airflow: Ensure you have Apache Airflow installed. Install Airflow.

Installation

1. Clone the repository:

git clone https://github.com/MFurmanczyk/wh-sales.git
cd wh-sales

2. Build the project:

  ./gradlew shadowJar

3. Setup Apache Airflow:

cd airflowdocker-compose up

4. Running the ETL
   Start the Apache Airflow web server and scheduler:

docker-compose up -d

5. Move dag.py to Airflow's DAGs folder.

Access the Airflow UI at http://localhost:8080 and trigger the ETL DAG (sales_dag).

License

This project is licensed under the MIT License - see the LICENSE file for details.