mirror of https://github.com/frappe/gunicorn.git synced 2026-07-01 18:21:30 +08:00

Ankush Menat 6d09a2e912 Merge upstream gunicorn 26.0.0 into Frappe fork

Reconcile the Frappe-specific gthread changes with upstream's major
26.0.0 release (HTTP/2, ASGI, lock-free main-thread event loop).

Conflict resolution:
- gunicorn/workers/gthread.py: reimplemented our features on top of
  upstream's rewritten event loop instead of merging line-by-line:
  - Per-request timeout + faulthandler traceback dump, tracked via an
    in-flight future set and keyed on exec_start_time so client-wait
    time (upstream's _DEFER path) is excluded and long-lived HTTP/2
    connections are exempt.
  - Adaptive slow/fast-lane queueing: the request-line peek/classify
    now reuses upstream's pending_conns poller-park mechanism, then
    dispatches to a fast or slow ThreadPoolExecutor.
  - Dropped our eventfd-based shutdown wakeup in favour of upstream's
    PollableMethodQueue.defer(), which already wakes the poller on
    SIGTERM.
- gunicorn/config.py: kept both new validators; our EnableAdaptiveQueueing
  and SlowRequestThreshold settings merged cleanly.
- README.rst was removed upstream (converted to README.md); ported an
  updated Fork Information section into README.md.
- tox.ini / .github/workflows/tox.yml: took upstream's Python matrix
  (it already drops the EoL versions our fork had pruned).

All gthread + routing tests pass (91), full suite 1976 passed / 263
skipped.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

2026-05-30 12:28:45 +05:30

4.5 KiB

Raw Blame History

Fork Information

This is a fork of gunicorn with the following Frappe-specific changes to the gthread worker:

Request timeout for gthread — upstream has no per-request timeout; a slow request can otherwise tie up a thread indefinitely. See https://github.com/frappe/gunicorn/pull/1
Traceback dump on timeout — when a request times out we faulthandler.dump_traceback() before exiting, to aid debugging. See https://github.com/frappe/gunicorn/pull/6
Adaptive slow/fast-lane queueing — when enable_adaptive_queueing is set, the configured threads are split into a fast lane and a slow lane so slow requests cannot starve fast ones. Controlled by --enable-adaptive-queueing / GUNICORN_ENABLE_ADAPTIVE_QUEUEING and --slow-request-threshold. See gunicorn/workers/gthread_routing.py.
Higher selector timeout — small optional throughput improvement. See https://github.com/frappe/gunicorn/pull/2

Note to anyone upgrading / adding changes:

Pull upstream changes.
Reapply or reconcile our changes (the work lives in gunicorn/workers/gthread.py, gunicorn/workers/gthread_routing.py and gunicorn/config.py).
Update the commit/ref in frappe/frappe pyproject.toml.
Keep this section up to date. Keep changes small and to the point.

TODO:

Plan a cleaner solution using custom workers. All current changes could live in a custom worker, but that carries the same awkwardness of maintaining a fork since we want to "override" rather than "extend" core behaviour.

Gunicorn

Gunicorn is maintained by volunteers. If it powers your production, please consider supporting us:

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 v25: Per-app worker allocation for dirty arbiters, HTTP/2 support (beta)!

Quick Start

pip install gunicorn
gunicorn myapp:app --workers 4

For ASGI applications (FastAPI, Starlette):

gunicorn myapp:app --worker-class asgi

Features

WSGI support for Django, Flask, Pyramid, and any WSGI framework
ASGI support for FastAPI, Starlette, Quart
HTTP/2 support (beta) with multiplexed streams
Dirty Arbiters (beta) for heavy workloads (ML models, long-running tasks)
uWSGI binary protocol for nginx integration
Multiple worker types: sync, gthread, gevent, asgi
Graceful worker process management
Compatible with Python 3.9+

Documentation

Full documentation at https://gunicorn.org

Community

Report bugs on GitHub Issues
Chat in #gunicorn on Libera.chat
See CONTRIBUTING.md for contribution guidelines

Support

Powering Python apps since 2010. Support continued development.

License

Gunicorn is released under the MIT License. See the LICENSE file for details.

4.5 KiB Raw Blame History