Adaptive Fault-Resistant Systems

摘要: Abstract : The combined effects of faults and resource failures, wide swings in service demand, situation-dependent user requirements, stress a computer's ability to satisfy its expectations. This is an especially significant problem distributed system that employ unreliable communications, whose components may operate different perhaps harsh physical data usage environments. One the goals for adaptive design allow flexible use available resources cover much wider range kinds environmental variables than could be covered by fixed, worst-case design. research focused on tasks of: (1) developing theory fault- resistant systems general principles architectural design; (2) specific techniques; (3) demonstrating designs. Three mechanisms were investigated: Adaptive Distributed Recovery Blocks (ADRBs), multiple-mode scheme error detection recovery, useful both hardware software faults; fault tolerance hybrid faults, economical technique tolerating simple complex types; thread integrity, detecting repairing breaks operating environments using Alpha programming model. systems, blocks, Fault tolerance, Anomaly management.