An algorithm to generate FSTs for the reliability analysis of distributed systems

The authors present a unified algorithm to efficiently generate file spanning trees (FSTs) and compute the distributed program reliability (DPR) and distributed system reliability (DSR). The algorithm is based on the concept of cutting one particular link in the distributed processing system (DPS) graph each time for generating different combination subgraphs. This approach guarantees no redundant FSTs to be generated during the subgraphs generating process. The reliability computation of the DPS is only needed to sum up the probability of the FSTs. A DPS reliability analysis tool based on the algorithm is implemented and is used to analyze several DPS examples for comparisons.<
>