AIRFLOW

Apache Airflow: Orchestration and First Look

Welcome to the first tutorial in the Airflow series. Before writing any code, we start with why cron jobs fall short for real data pipelines, what orchestration solves, and how Apache Airflow fits in. You'll get familiar with the core concepts and everything after this builds on them.

Orchestration

An orchestrator runs the steps of a data pipeline in the right order, tracks their status, and handles failures. Without one, a pipeline is a set of scripts you have to babysit.

Cron vs orchestration:

  • Cron job – schedules a script at a fixed time. Runs blindly. No visibility, no dependency tracking, no automatic recovery.
  • Orchestration – manages task order, dependencies, retries, and failure handling. Provides logs and UI for every run.

Think of the orchestrator as a conductor in an orchestra. Every task knows its own part, but the conductor decides who plays when and reacts if something goes wrong.

What a DAG Is

A DAG (Directed Acyclic Graph) is the structure that represents a pipeline.

  • Directed – tasks flow in one direction, upstream to downstream.
  • Acyclic – no loops. A task cannot depend on itself directly or through a cycle.
  • Graph – a set of connected tasks.

A pipeline that fetches data from an API, validates it, loads it into a database, and sends a report is a DAG with four tasks connected in sequence.

Airflow

Airflow is the most widely used orchestrator in data engineering. It was built at Airbnb in 2014, open sourced, and is now the industry standard. Pipelines are defined in Python.

Three core concepts:

  • DAG – the entire workflow, defined in one Python file.
  • Task – a single unit of work. One box in the DAG.
  • Scheduler – the component that decides when each DAG runs and dispatches its tasks.

Every run of a DAG has a timestamp, a duration, and a status. Clicking a run in the UI shows which tasks succeeded, which failed, and how long each one took.

Starting the Airflow Environment

Scroll to the bottom of the tutorial page and press Start Airflow. Once the environment is ready, an environment panel appears.

Interaction options:

  • VS Code – for editing DAG files with an IDE.
  • Terminal – for command-line access.

The Airflow UI link opens the web interface. Sign in with username and password both set to airflow.

The Airflow Homepage

Once signed in, the homepage displays:

  • DAG stats – failed DAGs, import errors, running DAGs, active DAGs. All start at zero.
  • Health – shows whether the scheduler, triggerer, and other components are running correctly.
  • Recent history – lists every DAG run in the last 24 hours with status and timestamps.

What's next

Now go and try this out in a live environment — boot a fresh cluster and play with the manifests above.

Start Airflow
Spec 4 CPU / 8 GiB ·Disk 25 GiB
Sign in to launch this environment
Required 1 VM · 4 CPU · 8 GB · 25 GiB disk
Available 1 VM · 1 CPU · 2 GB · 10 GiB disk
Sign in
Up next in Apache Airflow Chapter 2 of 12

Airflow: Write and Run Your First DAG

Continue

Suggested tutorials

Keep the momentum going, pick a related topic.