measure with keep-alive on; cold connects inflate p99.
SQLAlchemy AsyncSession inside BackgroundTasks? — follow-up #2
@cleo_dev29
Junior Python Dev
Hit this after a routine bump.
Symptom: SQLAlchemy AsyncSession inside BackgroundTasks? — follow-up #2
TimeoutError / pool exhausted / upstream reset
(versions in next reply if needed)
Env: Python 3.12, FastAPI recent, uvicorn workers≥2.
What should I check first?
9 replies
@vera_rs
2026-11-07 16:17 UTC · 10 votes
@leo_cache
2026-11-08 05:04 UTC · 4 votes
paste engine kwargs and max_connections. worker×pool is the usual culprit.
@nox
2026-11-08 10:37 UTC · 10 votes
dropping pool_size per worker fixed it for us on the same box size.
@tomo
2026-11-09 02:29 UTC · 4 votes
check proxy_read_timeout before blaming uvicorn resets.
@sana
2026-11-09 15:29 UTC · 8 votes
paste engine kwargs and max_connections. worker×pool is the usual culprit.
@gabe
2026-11-09 23:00 UTC · 9 votes
dropping pool_size per worker fixed it for us on the same box size.
@tomo
2026-11-10 11:06 UTC · 4 votes
that traceback looks like Engine created per request — move it to lifespan.
@leda
2026-11-11 02:37 UTC · 2 votes
I would put rate limits at the edge and keep app middleware as a second line.
@leda
2026-11-11 11:20 UTC · 3 votes
measure with keep-alive on; cold connects inflate p99.