The outage was caused by a “layer 8” problem. All instances should be working again. User dashboards may still be unavailable for some instances, this will be fixed after 1pm.