Abstract: Systems and methods for implementing a memory-efficient fault tolerant computing system are provided. A generic backup process may provide fault tolerance to multiple working processes. The backup process need not include a copy of the code segments executed by the working processes, providing very large savings in memory needed to implement the fault tolerant system. Alternatively, multiple backup processes provide fault tolerance but need not include duplicated code segments for the working processes they support.

Abstract: A mechanism for guaranteeing a coherent repository, including a list of requested files that are to be committed to the repository or to be processed (e.g. read) by an operation, and a list of “protected” files that are currently being processed. Either all of or none of the requested files are committed or are processed. Therefore, if any one of the requested files is in the protected-files list, then the invention waits until none of the requested files is protected before committing the requested files to the repository or before processing the requested files. An abnormal operation may hold up the protected files in the protected-files list for an unreasonably-long period of time. The invention thus provides an alarm mechanism, which scans the protected-files list to detect whether an abnormal operation exists. If such an operation is found, then the invention alerts an alarm agent and replaces the inconsistent files caused by the abnormal operation.