`ValueError: signal only works in main thread of the main interpreter`

[Avasam]

Describe the bug
When trying to enqueue a worker, I get the following error:

Traceback (most recent call last):
  File "/root/.cache/pypoetry/virtualenvs/backend-MATOk_fk-py3.12/lib/python3.12/site-packages/scheduler/helpers/queues/queue_logic.py", line 48, in perform_job
    job_model.success_callback(job_model, connection, result)  # type: ignore
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/root/.cache/pypoetry/virtualenvs/backend-MATOk_fk-py3.12/lib/python3.12/site-packages/scheduler/helpers/callback.py", line 36, in __call__
    with SCHEDULER_CONFIG.DEATH_PENALTY_CLASS(self.timeout, JobTimeoutException):
         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/root/.cache/pypoetry/virtualenvs/backend-MATOk_fk-py3.12/lib/python3.12/site-packages/scheduler/timeouts.py", line 32, in __enter__
    self.setup_death_penalty()
  File "/root/.cache/pypoetry/virtualenvs/backend-MATOk_fk-py3.12/lib/python3.12/site-packages/scheduler/timeouts.py", line 62, in setup_death_penalty
    signal.signal(signal.SIGALRM, self.handle_death_penalty)
  File "/usr/local/lib/python3.12/signal.py", line 58, in signal
    handler = _signal.signal(_enum_to_int(signalnum), _enum_to_int(handler))
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ValueError: signal only works in main thread of the main interpreter



During handling of the above exception, another exception occurred:



Traceback (most recent call last):
  File "/root/.cache/pypoetry/virtualenvs/backend-MATOk_fk-py3.12/lib/python3.12/site-packages/scheduler/helpers/queues/queue_logic.py", line 289, in run_sync
    result = perform_job(job, self.connection)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/root/.cache/pypoetry/virtualenvs/backend-MATOk_fk-py3.12/lib/python3.12/site-packages/scheduler/helpers/queues/queue_logic.py", line 52, in perform_job
    job_model.failure_callback(job_model, connection, *sys.exc_info())  # type: ignore
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/root/.cache/pypoetry/virtualenvs/backend-MATOk_fk-py3.12/lib/python3.12/site-packages/scheduler/helpers/callback.py", line 36, in __call__
    with SCHEDULER_CONFIG.DEATH_PENALTY_CLASS(self.timeout, JobTimeoutException):
         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/root/.cache/pypoetry/virtualenvs/backend-MATOk_fk-py3.12/lib/python3.12/site-packages/scheduler/timeouts.py", line 32, in __enter__
    self.setup_death_penalty()
  File "/root/.cache/pypoetry/virtualenvs/backend-MATOk_fk-py3.12/lib/python3.12/site-packages/scheduler/timeouts.py", line 62, in setup_death_penalty
    signal.signal(signal.SIGALRM, self.handle_death_penalty)
  File "/usr/local/lib/python3.12/signal.py", line 58, in signal
    handler = _signal.signal(_enum_to_int(signalnum), _enum_to_int(handler))
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ValueError: signal only works in main thread of the main interpreter

To Reproduce
Steps to reproduce the behavior:

1. I created a method sync_shopify_event_orders(event_id: int) under api/tasks.py with some logs inside.
2. I created a new task in the Django Admin Panel with the following configs:
   
3. Go to queues
4. Press enqueue

Expected behavior
No error

Screenshots
(see above)

Desktop (please complete the following information):

• OS: Docker python:3.12-slim
• python version: 3.12
• django version: 5.0.12
• pyproject dependencies:

dependencies = [
  "django ~=5.0.6",
  "django-allauth[socialaccount] ~=0.62.1",
  "django-anymail ~=10.3",
  "django-cors-headers ~=4.3.1",
  "django-environ ~=0.11.2",
  "django-tasks-scheduler ~=4.0.5",
  "django-widget-tweaks ~=1.5.0",
  "djangorestframework ~=3.15.1",
  "djangorestframework-camel-case ~=1.4.2",
  "gunicorn[gevent] ~=22.0.0",
  "pillow ~=10.4.0",
  "psycogreen ~=1.0.2",
  "psycopg2-binary ~=2.9.9",
  "uvicorn[standard] ~=0.29.0",
  "zygoat-django ~=1.0.1",
]

Additional context
This is my Django server start command in the dockerfile:

CMD poetry run ./manage.py wait_for_db \
  && poetry run ./manage.py migrate \
  && poetry run ./manage.py migrate --database=frontend_data \
  && poetry run ./manage.py createcachetable \
  && poetry run uvicorn --host 0.0.0.0 --port 3001 --reload backend.asgi:application

If I run poetry run ./manage.py scheduler_worker directly in the Docker, I see my task's own logs, followed by WARNING:scheduler:Job default:4:20250630180506946254 failed with exception: signal only works in main thread of the main interpreter.

And it runs every 10 seconds.

[cunla]

Can you provide me with your QUEUE configuration?
I suspect it is configured to be synced, and that's the root cause, but would like to confirm before I proceed

[cunla]

These errors are from the worker?

[Avasam]

Can you provide me with your QUEUE configuration?

Certainly! Forgot to share that as well.

# Tasks Scheduler
# https://django-tasks-scheduler.readthedocs.io/en/stable/configuration/
SCHEDULER_QUEUES: dict[str, QueueConfiguration] = {
    # Reusing our Django Redis cache container to avoid adding yet another image
    "default": QueueConfiguration(HOST="cache", PORT=6379, DB=0, ASYNC=False),
}

These errors are from the worker?

The stacktrace (code block + screenshot at the top of the issue) are what's visible from the /scheduler/jobs/default:4:20250630182323467974/ page

[cunla]

Ok, as suspected, it is happening because of ASYNC=False on the queue.

[Avasam]

Ok, just tried with ASYNC=True and I see my job being queued and becoming active. I think it's working with it. I don't remember exactly what it does or why I wouldn't want it. Probably just an artefact of copying the simplest example from the doc.

[cunla]

I still would like to fix it when ASYNC=False, but thanks for clarifying