We need to pull the plug on the dev and stage boxes with production data and see how well we do with disaster recovery 1) Soft reboot 2) Hard reboot 3) Server failover 4) Full server rebuild