We have identified and fixed the issue. All services are stable again, and we are continuing to monitor them.