Fault-Tolerant and Real-Time Distributed Computing

摘要

Progress was made on a number of problems in the areas of fault tolerant and realtime computing. Programming logics were investigated for reasoning about distributed programs that must satisfy real time constraints, must interact with a continuous physical environment, and whose correctness depends on properties of schedulers and degree of resource contention. A new approach to fault tolerance, based on a virtual machine monitor was developed. It provides fault tolerance without requiring modifications to hardware or software. Finally, software to support mobile network agents was developed and released. Algorithms to implement agent fault-tolerance were developed.