How do you run background tasks with Celery?

Question

Accepted Answer

**Celery** is the standard tool for running **background and scheduled tasks** in Django — work that shouldn't block the HTTP request (sending emails, processing files, calling slow APIs) or that runs on a schedule. Tasks are queued and processed asynchronously by separate **worker** processes.

## The problem Celery solves

```python
# ❌ doing slow work inline blocks the response — the user waits, risks timeouts
def register(request):
    user = create_user(request.POST)
    send_welcome_email(user)           # slow — the user waits for the email to send
    generate_report(user)               # even slower
    return redirect("home")             # finally responds after all that
```

Doing slow work inline makes users wait and risks request timeouts. Celery offloads it to background workers so the request returns immediately.

## Defining and calling a Celery task

```python
# tasks.py — define a task
from celery import shared_task

@shared_task
def send_welcome_email(user_id):
    user = User.objects.get(id=user_id)
    send_email(user.email, "Welcome!")    # runs in a WORKER, not the web process

# in a view — QUEUE the task instead of running it inline
def register(request):
    user = create_user(request.POST)
    send_welcome_email.delay(user.id)     # .delay() queues it → returns IMMEDIATELY
    return redirect("home")                # user gets a fast response
```

The `@shared_task` decorator defines a task; `.delay()` (or `.apply_async()`) queues it for a worker to process asynchronously — the view doesn't wait.

## The architecture

```text
Django (web) --.delay()--> BROKER (Redis/RabbitMQ) --> CELERY WORKER (processes the task)
                              (the message queue)         (separate process/machine)
```

```text
Components:
  ✓ Broker (Redis or RabbitMQ) — holds the queue of pending tasks
  ✓ Celery worker(s) — separate processes that pull and execute tasks
  ✓ Result backend (optional) — stores task results/status
Run workers separately: `celery -A myproject worker`
```

## Scheduled / periodic tasks (Celery Beat)

```python
# run tasks on a schedule (like cron) via Celery Beat
CELERY_BEAT_SCHEDULE = {
    "daily-report": {
        "task": "myapp.tasks.generate_daily_report",
        "schedule": crontab(hour=0, minute=0),     # every midnight
    },
}
```

## Reliability features

```python
@shared_task(bind=True, max_retries=3, default_retry_delay=60)
def sync_data(self):
    try:
        do_sync()
    except Exception as exc:
        raise self.retry(exc=exc)             # automatic retries on failure
```

```text
✓ Retries with backoff (resilience against transient failures)
✓ Task persistence (queued tasks survive restarts via the broker)
✓ Scheduling (Celery Beat — periodic/cron-like tasks)
✓ Rate limiting, concurrency control, task routing/priorities
✓ Workers scale independently of web servers
```

## Why it matters

Background task processing with Celery is essential knowledge for building responsive, scalable Django applications, since many real-world needs require work that shouldn't happen inline during a request.

Understanding it is important because the core problem — offloading slow or deferred work (sending emails, processing images/files, generating reports, calling slow external APIs, data syncing) from the request-response cycle — is ubiquitous in production applications.

Doing such work inline makes users wait, risks request timeouts, and loses the work if it fails, whereas **Celery** lets you respond to the user immediately while separate **workers** process the work asynchronously and reliably.

Beyond simple async offloading, Celery provides critical production capabilities: **scheduled/periodic tasks** (via Celery Beat, replacing cron for application-level scheduling — daily reports, cleanup jobs, recurring syncs), **automatic retries with backoff** (resilience against transient failures), **task persistence** (queued tasks survive restarts via the broker, so work isn't lost), and **independent scaling** of workers from web servers.

Understanding Celery's architecture (the broker like Redis/RabbitMQ holding the queue, workers processing tasks, the result backend), how to define and queue tasks (`@shared_task`, `.delay()`), schedule periodic work, and configure reliability features is valuable senior-level knowledge for building production Django systems that handle heavy or deferred work gracefully.

Because background processing is a common, important requirement for responsive, robust applications — and Celery is the de facto standard solution in the Django ecosystem — mastering it is a practical, frequently-relevant skill that reflects the ability to architect real-world Django applications beyond simple synchronous request handling.