“On the Performance of Byzantine Fault-Tolerant MapReduce”
From Navigators
(Difference between revisions)
(Created page with "{{Publication |type=article |document=Document for Publication-PedroCosta 2013.pdf |title=On the Performance of Byzantine Fault-Tolerant MapReduce |author=Pedro Costa and Marcelo...") |
|||
| (One intermediate revision not shown) | |||
| Line 3: | Line 3: | ||
|document=Document for Publication-PedroCosta 2013.pdf | |document=Document for Publication-PedroCosta 2013.pdf | ||
|title=On the Performance of Byzantine Fault-Tolerant MapReduce | |title=On the Performance of Byzantine Fault-Tolerant MapReduce | ||
| - | |author=Pedro Costa | + | |author=Pedro Costa, Marcelo Pasin, Alysson Bessani, Miguel Correia |
| + | |Project=Project:TCLOUDS, Project:RC-Clouds, | ||
| + | |ResearchLine=Fault and Intrusion Tolerance in Open Distributed Systems (FIT) | ||
|year=2013 | |year=2013 | ||
|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: non-speculative and speculative. | |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: non-speculative 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. | 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. | ||
| - | |||
|journal=IEEE Transactions on Dependable and Secure Computing | |journal=IEEE Transactions on Dependable and Secure Computing | ||
}} | }} | ||
Latest revision as of 18:33, 31 January 2013
Pedro Costa, Marcelo Pasin, Alysson Bessani, Miguel Correia
IEEE Transactions on Dependable and Secure Computing, 2013.
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: non-speculative 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.
Download paper
Download On the Performance of Byzantine Fault-Tolerant MapReduce
Export citation
Project(s): Project:TCLOUDS, Project:RC-Clouds
Research line(s): Fault and Intrusion Tolerance in Open Distributed Systems (FIT)
