Celery Best Practices

No.2: Use more Queues (ie. not just the default one)

.. The way to solve the issue above is to have taskA in one queue, and taskB in another and then assign x workers to process Q1 and all the other workers to process the more intensive Q2 as it has more tasks coming in. This way you can still make sure that taskB gets enough workers all the while maintaining a few priority workers that just need to process taskA when one comes in without making it wait to long on processing.