Two node reboots were needed to resolve an internal monitoring issue. We apologize for the inconvenience caused by this brief outage.