Celery任务集中子任务的自定义路由

Question

Is it possible to set subtask parameters "queue" and "routing_key" not in apply_async().

It is necessary in Taskset that each subtask went to custom queue and routing_key.

Looks like:

tasks = []
for item in items:   
    tasks.append(task_name.s((params).<method for set custom queue and routing_key>))
job = TaskSet(tasks=tasks)

Answer 1

It is not too clear what is the question. If you are asking if you can give routing_key and queue arguments to subtasks you can if you use subtask() instead of s() .

From subtask doc:

options – Additional options to Task.apply_async().

Answer 2

Yes, you can use any of the apply_sync arguments in the subtask call.

See the "Calling API" in the docs