Abstract: A data high-availability subsystem-DHAM（data high availability module）-for a distributed file system is presented in this paper. It is different from the state-of-the-art schemes in which the data high availability implementation is embedded in the file system itself. The main idea is to extract the high availability implementation from the distributed file system and make it a separate module, and then keep simple interface between them. Two benefits are achieved from this scheme: the first, it makes the file system implementation more modular; the second, it is more convenient to make a special study of fault-tolerance of distributed file systems. DHAM has been successfully used in Clufs, a distributed file system implemented by the authors. Finally, the function and performance test results and the theoretical analysis are presented.