Introduction
A high-performance, asynchronous, ready for production job scheduling framework.
TaskShed provides a simple API to schedule your Python coroutines for later execution. You can run tasks just once or on a recurring interval. The scheduler is dynamic, allowing you to add, update or remove tasks on the fly. Furthermore, by connecting to a persistent datastore, TaskShed ensures your tasks survive restarts and automatically catches up on any executions that were missed while the system was offline.
The key features are:
- Fast: TaskShed has an extremely low latency, overhead and can execute several thousands tasks a second.
- Distributed: TaskShed has the capacity to spawn several workers and schedules across many machines, while also providing support for monolinth architectures.
- Persistant: Tasks are stored in database, meaning that tasks won't get dropped on shutdown. TaskShed currently supports Redis and MySQL.
- Easy: TaskShed is straightforward to run.
Installation
The base version of TaskShed has no additional dependencies, though needs a driver for task persistance. This can be installed by running:
Quickstart
Create a file main.py with:
from datetime import datetime, timedelta
from taskshed.datastores import InMemoryDataStore
from taskshed.schedulers import AsyncScheduler
from taskshed.workers import EventDrivenWorker
async def say_hello(name: str):
print(f"Hello, {name}!")
datastore = InMemoryDataStore()
worker = EventDrivenWorker(callback_map={"say_hello": say_hello}, datastore=datastore)
scheduler = AsyncScheduler(datastore=datastore, worker=worker)
async def main():
await scheduler.start()
await worker.start()
await scheduler.add_task(
callback="say_hello",
run_at=datetime.now() + timedelta(seconds=3),
kwargs={"name": "World"},
)
if __name__ == "__main__":
import asyncio
loop = asyncio.new_event_loop()
loop.create_task(main())
loop.run_forever()
Then run it with.