Adding fault-tolerance to a hierarchical DRE system

Dynamic resource management is a crucial part of the infrastructure for emerging mission-critical distributed real-time embedded system. Because of this, the resource manager must be fault-tolerant, with nearly continuous operation. This paper describes an ongoing effort to develop a fault-tolerant multilayer dynamic resource management capability and the challenges we have encountered, including multi-tiered structure, rapid recovery, the characteristics of component middleware, and the co-existence of replicated and non-replicated elements. While some of these have been investigated before, this work exhibits all of these characteristics simultaneously, presenting a significant fault-tolerance research challenge.