One of our EU clusters experienced a temporary disconnection from it's broker and accumulated a large number of tasks resulting in a temporary spike of 429s for many users. The issue is now resolving and a fix is being implemented to prevent abnormal task accumulation.