One our our backend servers had an incident and temporarily went offline. The incident was resolved within 20 minutes, with a major outage for our auth services and minor service degradation to other services as their instances moved to other backend servers.