~/talk

Free Threaded Django with Micah Lyle

Talkat DjangoCon US 2025 · on Wednesday, 2025-09-10· at 14:55· Sauganash Ballroom· [django, devops-and-delivery]
slides:none on file

Demo runs show runserver (one thread per request) using a pure Python hasher cuts developer latency and records about 40% faster on a laptop, while Gunicorn with 16 process workers gives predictable throughput. Gunicorn gthread (1 process, 16 threads) matches throughput with much lower memory use. Graian (ASGI Rust server) sometimes created multiple ninjas and crashes, exposing thread races. Recommendation: sync Django stays with multiprocessing unless memory is tight; async CPU‑bound apps benefit from free‑threaded mode.

AI-written summary · about

  1. 0:00Free-threaded Django intro
  2. 1:25GIL vs free threading demo
  3. 9:50OS and threads crash course
  4. 27:08Django thread safety check
  5. 32:25Async + threading in Django
  6. 36:19Free threading recommendations
  7. 39:11Q&A on sub-interpreters
~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.