Abstract: The fault tolerant parallel algorithm (FTPA) is an application-level technique for tolerating hardware failures.FTPA achieves fast failure recovery making use of parallel recomputing. How to deal with system failures is a concern in the design of FTPA. Thus, evaluating the performance of FTPA under system failures is necessary. In this study,we presented the performance metrics to evaluate the performance of FTPA and a model to predict the application completion time under system failures. Then, the influence of program section executing time, checkpointing cost, failureate, and speedup of parallel recomputing on the performance of FTPA were evaluated.

Key words: Fault-tolerant parallel algorithm, Application completion time, Speedup, Efficiency