[dagi3d]

eg. That request might take more time than what a http request should. Also, by having workers that consume specific queues/topics, it's easier to scale them with more granularity