The Community for Technology Leaders
Green Image
Issue No. 05 - Sept.-Oct. (2013 vol. 10)
ISSN: 1545-5971
pp: 301-313
Alysson Neves Bessani , University of Lisboa, Lisboa
Miguel P. Correia , Technical University of Lisbon, Lisboa
Marcelo Pasin , University of Neuchatel, Neuchatel
Pedro Costa , Universidade de Lisboa, Lisboa
ABSTRACT
MapReduce is often used for critical data processing, e.g., in the context of scientific or financial simulation. However, there is evidence in the literature that there are arbitrary (or Byzantine) faults that may corrupt the results of MapReduce without being detected. We present a Byzantine fault-tolerant MapReduce framework that can run in two modes: nonspeculative and speculative. We thoroughly evaluate experimentally the performance of these two versions of the framework, showing that they use around twice more resources than Hadoop MapReduce, instead of the three times more of alternative solutions. We believe this cost is acceptable for many critical applications.
INDEX TERMS
Fault tolerance, Fault tolerant systems, Impedance matching, Computational modeling, Google, Computer crashes, Heart beat, Byzantine fault tolerance, Hadoop, MapReduce
CITATION
Alysson Neves Bessani, Miguel P. Correia, Marcelo Pasin, Pedro Costa, "On the Performance of Byzantine Fault-Tolerant MapReduce", IEEE Transactions on Dependable and Secure Computing, vol. 10, no. , pp. 301-313, Sept.-Oct. 2013, doi:10.1109/TDSC.2013.7
86 ms
(Ver )