We had an unscheduled downtime yesterday (21st of July) at 23:40 due to a power failure. The cluster was back in operation 1 hour 20 minutes later.