tarsq 0.1.14

A fast, production-ready task queue for Python

tarsq

A fast, production-ready task queue for Python backed by Redis.

Features

• Simple @task decorator to register handlers (sync and async)
• dispatch() to enqueue jobs from anywhere in your app
• status() to track job progress
• Automatic retries (up to 3) with exponential backoff
• Configurable worker count and task timeout
• Startup/shutdown lifecycle hooks
• Crashed worker auto-restart
• Stuck task recovery on startup

Requirements

• Python 3.11+
• Redis

Installation

pip install tarsq

Quick start

1. Define your tasks (myapp/tasks.py):

from tarsq import task

@task("send_email")
def send_email(ctx, payload):
    print(f"Sending email to {payload['to']}")

@task("resize_image")
async def resize_image(ctx, payload):
    print(f"Resizing image at {payload['url']}")

2. Configure the worker (myapp/worker_settings.py):

from tarsq import WorkerSettings

class MyWorkerSettings(WorkerSettings):
    app = "myapp.tasks"
    workers = 5
    timeout = 60

3. Start the worker:

tarsq --settings myapp.worker_settings.MyWorkerSettings

Or without a settings class:

tarsq --app myapp.tasks --workers 5 --timeout 60

4. Dispatch jobs from your application:

from tarsq import dispatch, status

job_id = dispatch("send_email", payload={"to": "[e-mail]"})

job = status(job_id)
print(job.status)  # queued | in_progress | completed | failed

Configuration

WorkerSettings

Attribute	Type	Default	Description
app	str	None	Dotted module path containing your @task handlers
workers	int	5	Number of concurrent worker threads
timeout	int	300	Max seconds a task can run before being killed
ctx	dict	{}	Shared context passed to every task handler
on_startup	callable	None	Called once before workers start (sync or async)
on_shutdown	callable	None	Called once after workers stop (sync or async)

Environment variables

Variable	Default	Description
REDIS_HOST	localhost	Redis host
REDIS_PORT	6379	Redis port
REDIS_PASSWORD	None	Redis password

Variables can also be set in a .env file in your project root.

Lifecycle hooks

from tarsq import WorkerSettings

async def on_startup(ctx):
    ctx["db"] = await create_db_pool()

async def on_shutdown(ctx):
    await ctx["db"].close()

class MyWorkerSettings(WorkerSettings):
    app = "myapp.tasks"
    on_startup = on_startup
    on_shutdown = on_shutdown

The ctx dict is passed into every task handler, making it easy to share connections across tasks.

Job status

from tarsq import status, TaskStatusEnum

job = status(job_id)

job.job_id      # UUID string
job.task        # task name
job.status      # TaskStatusEnum: queued | in_progress | completed | failed
job.retries     # number of retry attempts so far
job.created_at  # ISO 8601 timestamp
job.updated_at  # ISO 8601 timestamp of last status change

License

MIT