mirror of
https://github.com/frappe/gunicorn.git
synced 2026-07-03 11:11:30 +08:00
77222b8017
Introduce Dirty Arbiters - a separate process pool for executing long-running, blocking operations (AI model loading, heavy computation) without blocking HTTP workers. Inspired by Erlang's dirty schedulers. Key features: - Completely separate from HTTP workers - can be killed/restarted independently - Stateful - loaded resources persist in dirty worker memory - Message-passing IPC via Unix sockets with JSON serialization - Explicit execute() API from HTTP workers - Asyncio-based for clean concurrent handling Architecture: - DirtyArbiter: manages the dirty worker pool, routes requests - DirtyWorker: executes functions, maintains state, handles requests - DirtyClient: sync/async API for HTTP workers to call dirty apps - DirtyProtocol: length-prefixed JSON messages over Unix sockets - DirtyApp: base class for dirty applications Configuration options: - dirty_apps: list of import paths for dirty applications - dirty_workers: number of dirty workers (default: 0) - dirty_timeout: task timeout in seconds (default: 300) - dirty_graceful_timeout: shutdown timeout (default: 30) Lifecycle hooks: - on_dirty_starting(arbiter) - dirty_post_fork(arbiter, worker) - dirty_worker_init(worker) - dirty_worker_exit(arbiter, worker) Includes comprehensive test suite with 164 tests covering: - Protocol encoding/decoding - Worker and arbiter lifecycle - Client sync/async APIs - Signal handling - Error handling and timeouts - Integration tests
Gunicorn
========
.. image:: https://img.shields.io/pypi/v/gunicorn.svg?style=flat
:alt: PyPI version
:target: https://pypi.python.org/pypi/gunicorn
.. image:: https://img.shields.io/pypi/pyversions/gunicorn.svg
:alt: Supported Python versions
:target: https://pypi.python.org/pypi/gunicorn
.. image:: https://github.com/benoitc/gunicorn/actions/workflows/tox.yml/badge.svg
:alt: Build Status
:target: https://github.com/benoitc/gunicorn/actions/workflows/tox.yml
Gunicorn 'Green Unicorn' is a Python WSGI HTTP Server for UNIX. It's a pre-fork
worker model ported from Ruby's Unicorn_ project. The Gunicorn server is broadly
compatible with various web frameworks, simply implemented, light on server
resource usage, and fairly speedy.
**New in v24**: Native ASGI support (beta) for async frameworks like FastAPI!
Quick Start
-----------
.. code-block:: bash
pip install gunicorn
gunicorn myapp:app --workers 4
For ASGI applications (FastAPI, Starlette):
.. code-block:: bash
gunicorn myapp:app --worker-class asgi
Features
--------
- WSGI support for Django, Flask, Pyramid, and any WSGI framework
- **ASGI support** (beta) for FastAPI, Starlette, Quart
- uWSGI binary protocol for nginx integration
- Multiple worker types: sync, gthread, gevent, eventlet, asgi
- Graceful worker process management
- Compatible with Python 3.12+
Documentation
-------------
Full documentation at https://gunicorn.org
- `Quickstart <https://gunicorn.org/quickstart/>`_
- `Configuration <https://gunicorn.org/configure/>`_
- `Deployment <https://gunicorn.org/deploy/>`_
- `Settings Reference <https://gunicorn.org/reference/settings/>`_
Community
---------
- Report bugs on `GitHub Issues <https://github.com/benoitc/gunicorn/issues>`_
- Chat in `#gunicorn`_ on `Libera.chat`_
- See `CONTRIBUTING.md <CONTRIBUTING.md>`_ for contribution guidelines
Sponsors
--------
Gunicorn is maintained thanks to our sponsors. `Become a sponsor <https://github.com/sponsors/benoitc>`_.
.. Sponsor logos will appear here
License
-------
Gunicorn is released under the MIT License. See the LICENSE_ file for details.
.. _Unicorn: https://bogomips.org/unicorn/
.. _`#gunicorn`: https://web.libera.chat/?channels=#gunicorn
.. _`Libera.chat`: https://libera.chat/
.. _LICENSE: https://github.com/benoitc/gunicorn/blob/master/LICENSE