Recovery Schemes for High Availability and High Performance Distributed Real-Time Computing