Admin, agent, guest were currently unable to access our service for about 1.5 hours. The outage was caused by a failed infrastructure component that took 30 minutes to rebuild.