Airflow Fundamentals Overview
This Introduction to Airflow training course teaches software engineers and data engineers how to use Apache Airflow to orchestrate production-ready data pipelines. Students learn how to create pipelines via DAGs (Directed Acyclic Graphs), make pipelines predictable, and master their scheduling. In addition, participants learn how to abstract functionality, reuse components, and modularize and share data across tasks and pipelines with the TaskFlow API, Task Groups, and Branching. Finally, attendees learn how to scale Airflow within Celery, Kubernetes, and KEDA executors.
Throughout the course, students incrementally construct a single, real-world application.
Location and Pricing
Accelebrate offers instructor-led enterprise training for groups of 3 or more online or at your site. Most Accelebrate classes can be flexibly scheduled for your group, including delivery in half-day segments across a week or set of weeks. To receive a customized proposal and price quote for private corporate training on-site or online, please contact us.
In addition, some courses are available as live, instructor-led training from one of our partners.
Objectives
- Create production-ready data pipelines in Airflow
- Build pipelines in Airflow that are able to scale to hundreds of tasks
- Enforce modularization and reusability of Airflow tasks across projects
- Scale Airflow in Kubernetes
Prerequisites
All attendees should have basic Python knowledge or object-oriented programming experience.
Outline
Expand All | Collapse All
Introducing Apache Airflow
- What Airflow is and what does it solve?
- Airflow architecture
- How do we represent a Pipeline?
- Our first DAG
- Tasks, TaskFlow, and Operators
- First Pipeline
Mastering scheduling
- execution_date, start_date and schedule_interval
- Handling non-default schedule_intervals
- Playing with time
Abstracting functionality
- Using custom operators
- Creating TaskGroups vs subDAGs
- Sharing data with xCOMs
- Branching and Triggers
- Sensors and SmartSensors
Executors and Scaling Airflow
- Abandoning SQLite for PostgreSQL
- Executors: Debug, Local, Celery
- Concurrency and parallelism
- Concurrency with Celery
- Airflow in Kubernetes, the old and new ways
- KEDA and HA scheduler
- Deploying a highly availability fault-tolerant Airflow
Conclusion
Training Materials
All Airflow training students receive comprehensive courseware.
Software Requirements
- Python 3.5 or later
- Airflow 2.1 or later
