A dynamic fault-tolerant model for open distributed computing

Open distributed computer systems are some of the most successful structures ever designed for the computer community together with their undisputed benefits for users. However, their complexity has also introduced a few side-effects, most notably the unpredictable nature of the underlying environments and reconfiguration burdens imposed by environmental changes. Thus, to gain high level of system performance, a required level of reliability has to be maintained. In this paper, we propose a mechanism to analyze the underlying environmental faults and failures. This model provides an adaptable fault-tolerant approache in order to address unanticipated events and unpredictable hazards in distributed systems. Therefore, this model maintains the required reliability by analyzing the environment and selects the optimal replication strategy for existing conditions. This pragmatic and theoretically appealing approach is a part of the Juice system which supports adaptation properties for open distributed environments.