Skip to content

fix(celery): update worker concurrency defaults #1228

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
mihow merged 2 commits into from
Apr 15, 2026
Merged

fix(celery): update worker concurrency defaults #1228

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
44dd942
fix(celery): bump worker concurrency default to 16
mihow Apr 14, 2026
07a1c2b
chore(celery): lower default concurrency to 8, document prod override
mihow Apr 15, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions .envs/.local/.django
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,6 +19,11 @@ NATS_URL=nats://nats:4222
CELERY_FLOWER_USER=QSocnxapfMvzLqJXSsXtnEZqRkBtsmKT
CELERY_FLOWER_PASSWORD=BEQgmCtgyrFieKNoGTsux9YIye0I7P5Q7vEgfJD2C4jxmtHDetFaE2jhS7K7rxaf

# Celery worker concurrency (prefork pool size).
# Default in settings is 8. Lower to 2-4 on memory-constrained dev laptops —
# each prefork worker is a separate Python process with its own DB connection.
# CELERY_WORKER_CONCURRENCY=4

# RabbitMQ
CELERY_BROKER_URL=amqp://rabbituser:rabbitpass@rabbitmq:5672/
CELERY_RESULT_BACKEND=rpc:// # Use RabbitMQ for results backend
Expand Down
7 changes: 7 additions & 0 deletions .envs/.production/.django-example
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,13 @@ REDIS_URL=redis://redis:6379/0
# Celery
# ------------------------------------------------------------------------------

# Worker concurrency (prefork pool size). Settings default is 8.
# Recommended for production: 16 on an 8-core host — process_nats_pipeline_result
# and create_detection_images are DB/Redis-bound, so ~2x cpu_count roughly matches
# observed drain vs ingress on the antenna queue without saturating pgbouncer.
# Tune higher only after confirming postgres connection-pool headroom.
CELERY_WORKER_CONCURRENCY=16

# Flower
CELERY_FLOWER_USER=
CELERY_FLOWER_PASSWORD=
Expand Down
10 changes: 10 additions & 0 deletions config/settings/base.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -393,6 +393,16 @@ def _celery_result_backend_url(redis_url):
CELERY_WORKER_PREFETCH_MULTIPLIER = 1
CELERY_WORKER_ENABLE_PREFETCH_COUNT_REDUCTION = True

# Worker concurrency (prefork pool size)
# https://docs.celeryq.dev/en/stable/userguide/configuration.html#worker-concurrency
# Celery's own default when unset is os.cpu_count(), which on the production
# 8-core host produced an 8-process pool that could not keep up with the antenna
# queue's DB/Redis-bound tasks (process_nats_pipeline_result, create_detection_images).
# 8 is a conservative default that keeps local/staging/demo memory footprints
# reasonable (each prefork worker is a separate Python process with imports +
# DB connection). Production should override to 16 (see .envs/.production/.django-example).
CELERY_WORKER_CONCURRENCY = env.int("CELERY_WORKER_CONCURRENCY", default=8)

# Cancel & return to queue if connection is lost
# https://docs.celeryq.dev/en/latest/userguide/configuration.html#worker-cancel-long-running-tasks-on-connection-loss
CELERY_WORKER_CANCEL_LONG_RUNNING_TASKS_ON_CONNECTION_LOSS = True
Expand Down
Loading