Abstract
This article describes the automatic recovery algorithm and its implementation for the POOSS stable file system. The recoverable faults by our scheme include non-consecutive disk loss, single duplication loss, single duplication corruption, and duplication inconsistency. The recovery algorithm is implemented on the 100-node parallel machine (POOMA). One important feature of the recovery algorithm is that it is node-independent and can be executed simultaneously on different nodes, resulting in a highly parallel recovery procedure.
Cite
CITATION STYLE
Chengzheng, S., & Hertzberger, L. O. (1991). Automatic recovery of a parallel stable file system. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 503 LNCS, pp. 345–353). Springer Verlag. https://doi.org/10.1007/3-540-54132-2_65
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
