discussion 2026-11-20 1958 views 35 votes

Redis pipeline vs single GETs under 5k RPS

@tomo
Backend Engineer

Hit this after a routine bump.

Symptom: Redis pipeline vs single GETs under 5k RPS

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?

7 replies

@rina_lab79
2026-11-20 14:05 UTC · 9 votes

check proxy_read_timeout before blaming uvicorn resets.

@cleo_dev29
2026-11-20 14:31 UTC · 5 votes

marking solved after we pinned the dependency and re-smoked /docs.

@kai_sre
2026-11-20 21:52 UTC · 3 votes

paste engine kwargs and max_connections. worker×pool is the usual culprit.

@jon_async
2026-11-21 01:37 UTC · 10 votes

dropping pool_size per worker fixed it for us on the same box size.

@skye
2026-11-21 12:46 UTC · 4 votes

marking solved after we pinned the dependency and re-smoked /docs.

@mei
2026-11-21 23:22 UTC · 9 votes

dropping pool_size per worker fixed it for us on the same box size.

@sana
2026-11-22 13:22 UTC · 8 votes

orjson helped CPU a few percent; DB still the ceiling.

Reply

Login or register to reply.