selfhost.directory

The index of self-hostable projects — ranked, categorized and served with live updates.
Own your data and stop giving it to big tech.

Apache Airflow logo

Self-host Apache Airflow

3.4.0

Workflow orchestration for data pipelines and ML

ReleaseOpen SourceApache-2.0
Updatedyesterday
LanguagePython
Setup difficultyAdvanced
ThemeDark mode
Stars
47 415
Port
8080 (API server / web UI)
Docker
apache/airflow
DAGs list view in dark modeAsset dependencies graph viewGrid view spanning across timeGraph view of DAG dependenciesHome page with system health and statsBackfill interface for date rangesCode view for DAG source

What is Apache Airflow?

Apache Airflow lets you define workflows as Python code (DAGs) so they are versionable, testable, and collaborative. The scheduler runs tasks on workers while respecting dependencies. A modern web UI shows pipeline status, logs, and graphs with dark/light themes. It scales via Local, Celery, or Kubernetes executors and connects to cloud services, databases, and other tools through a large provider ecosystem. Common uses include data engineering pipelines, ML model training, and AI workload orchestration. The project is community-driven under the Apache Software Foundation.

Apache Airflow alternatives

All self-hosted alternatives to Azkaban →

Should you self-host Apache Airflow?

Apache Airflow scores 85/100 on selfhost.directory (how scoring works) and ranks #2 of 32 in Workflow Automation. It has 47 415 GitHub stars, +369 in the last 30 days.

It shipped 8 stable releases in the past year. The latest, 3.4.0, came out on 6 November 2026. New versions land roughly every 2 weeks.

Setup is advanced: plan for several services and some ongoing maintenance. An official Docker image is available. It is Apache-2.0-licensed and written mostly in Python.

Its closest alternative on selfhost.directory is n8n (92/100, 209 773 stars). It scores higher overall, so compare the two before you pick. See Apache Airflow vs n8n side by side.

Features

  • Audit Logging
  • Dynamic DAG Generation
  • Extensible Operators
  • Jinja Templating
  • Kerberos Authentication
  • Multiple Executors
  • Multi-tenancy
  • REST API
  • Rich Web UI
  • Secrets Management
  • XCom Cross-Task Communication

Apache Airflow README

Apache Airflow

Category Badges
License License
PyPI PyPI version PyPI - Python Version PyPI - Downloads
Containers Docker Pulls Docker Stars Artifact HUB
Community Contributors Slack Status Commit Activity LFX Health Score
Dev tools prek
Version Build Status
Main Tests AMD main Tests ARM main
3.x Tests AMD 3.2 Tests ARM 3.2

Apache Airflow releases & version history

@ Every self-hosted release, weekly

One Monday email with the week's stable releases across the directory.

FAQ

Why did the pause DAG toggle turn red?

Pausing or unpausing the DAG failed; the toggle reverts to previous state and turns red. Try again or check logs.

How do I stop the sync perms happening multiple times per webserver?

Set [fab] update_fab_perms configuration in airflow.cfg to False.

What does "MySQL Server has gone away" mean?

Connection pool keeping connections open too long. Set sql_alchemy_pool_recycle to invalidate connections after a specified time.

Why does my DAG version keep increasing?

DAG version inflation occurs when constructor arguments use dynamic values like datetime.now(), random values, or environment variables that change on each parse. Use fixed values and Jinja templates instead.

How to improve DAG performance?

Adjust parallelism, max_active_tasks_per_dag, max_active_runs_per_dag configurations. Use DAG-level max_active_tasks, max_active_runs, and task-level pool, priority_weight, queue settings.

More in Workflow Automation

Rate this project

Click to rate — no account needed.

0 comments

Posting as guest
Loading comments…