The incident was caused by an overloaded message queue causing resource contention with the application. Our engineers are investigating targeting alerting policies to catch this incident sooner as well as changes we can enact sooner to ensure customers are not affected by this issue in the future.