Generalized Evaluation Model with File Recovery Mechanism for Fault Latency

摘要

This article considers and extended evaluation model with file recovery for fault latency, analyzes it on more general assumptions, and discusses the optimal interval of archive copies. First, using archive copies, we evaluate the recovery overhead until the process restores a consistent file state just before a fault becomes latent. It is assumed that both the distribution of the time from the newest archive copy to a failure occurrence and the distribution of fault latency time are general.