At 5:46pm EST a misconfiguration in an infrastructure change was deployed that caused latency in some background event processing. The change was reverted at 6:14pm and background event processing recovered. At 6:28pm the system fully recovered.