The design and implementation of a reliable version of the distributed bitonic sorting algorithm using the application-oriented fault tolerance paradigm on a commercial multicomputer is described. Sorting assertions in general are discussed and the bitonic sort algorithm is introduced. Faulty behavior is discussed and a fault-tolerant parallel bitonic sort developed using this paradigm is presented. The error coverage and the response of the fault-tolerant algorithm to faulty behavior are presented. Both asymptotic complexity and the results of run-time experimental measurements on an Ncube multicomputer are given. The authors demonstrate that the application-oriented fault tolerance paradigm is applicable to problems of a noniterative nature
B. M. McMillin and L. M. Ni, "Reliable Distributed Sorting Through the Application-oriented Fault Tolerance Paradigm," Proceedings of the 9th International Conference on Distributed Computing Systems, 1989, Institute of Electrical and Electronics Engineers (IEEE), Jan 1989.
The definitive version is available at http://dx.doi.org/10.1109/ICDCS.1989.37983
9th International Conference on Distributed Computing Systems, 1989
Keywords and Phrases
Ncube Multicomputer; Application-Oriented Fault Tolerance Paradigm; Asymptotic Complexity; Commercial Multicomputer; Design; Distributed Processing; Error Coverage; Fault Tolerant Computing; Faulty Behaviour; Implementation; Multiprocessing Systems; Reliable Distributed Sorting; Sorting
Article - Conference proceedings
© 1989 Institute of Electrical and Electronics Engineers (IEEE), All rights reserved.