django-tasks-db-async

Async worker for django-tasks-db with OpenTelemetry instrumentation.

Table of Contents

• Installation
• Configuration
• Running the worker
• OpenTelemetry
• License

Installation

pip install django-tasks-db-async

Configuration

Add both django_tasks_db and django_tasks_db_async to INSTALLED_APPS, then configure the TASKS setting to use django_tasks_db.DatabaseBackend:

INSTALLED_APPS = [
    ...
    "django_tasks_db",
    "django_tasks_db_async",
]

TASKS = {
    "default": {
        "BACKEND": "django_tasks_db.DatabaseBackend",
        "QUEUES": ["default"],
    },
}

Run the database migrations to create the task result table:

python manage.py migrate

Running the worker

Start the worker with the db_async_worker management command:

python manage.py db_async_worker

Options

Examples

Process tasks from all queues with 4 concurrent workers:

python manage.py db_async_worker --queue-name='*' --concurrency=4

Process tasks from specific queues:

python manage.py db_async_worker --queue-name=default,emails,notifications

Run in batch mode — process outstanding tasks and exit (useful for cron jobs):

python manage.py db_async_worker --batch

Signals

The worker dispatches the standard task_started and task_finished Django signals around each task execution, compatible with both django.tasks (Django 6.0+) and the standalone django-tasks package.

OpenTelemetry

The worker records a span for each task execution using the OpenTelemetry messaging semantic conventions. No additional configuration is required — the worker picks up whatever TracerProvider is active in the process.

Each span includes the following attributes:

Failed tasks set the span status to ERROR and record the exception as a span event.

License

django-tasks-db-async is distributed under the terms of the MIT license.