Retry

Retry extension enqueues task again if it raised an exception. 

from aiotaskqueue import Configuration, task
from aiotaskqueue.extensions.builtin import Retry, RetryExtension
from aiotaskqueue.serialization.msgspec import MsgSpecSerializer


@task(
    name="name",
    markers=[Retry(max_retries=3)],  # (1)!
)
async def some_task() -> None:
    pass


configuration = Configuration(
    default_serialization_backend=MsgSpecSerializer(),
    extensions=[RetryExtension()],  # (2)!
)

  1. You need to add the Retry marker and configure amount of retries
  2. Don't forget to add the extension into your configuration

Scheduling

On retry task is simply added to the queue again, so the following applies:

  • Task may be scheduled on different node
  • Task won't be retried immediately if there are any tasks in front of it in the queue