Enhanced fault-tolerance through Byzantine failure detection

Rida Bazzi; Maurice Herlihy

doi:10.1007/978-3-642-10877-8_12

Enhanced fault-tolerance through Byzantine failure detection

Rida Bazzi, Maurice Herlihy

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

3 Scopus citations

Abstract

We consider a variant of the Byzantine failure model in which Byzantine processes are eventually detected and silenced, and investigate the fault-tolerance of the classical broadcast and agreement problems. We show that if all Byzantine processes are eventually detected, then it is possible to solve the broadcast problem in the presence of any number of Byzantine processes. If only a fraction of the Byzantine processes can be detected, then we show that it is possible to solve consensus (and broadcast) if the total number of processes is N ≥ 2f + 3F + 1, where f is the number of Byzantine processes that are eventually detected and F is the number of those that are never detected. We show that 2f + 3F + 1 is a lower bound to solve the consensus and broadcast problems.

Original language	English (US)
Title of host publication	Principles of Distributed Systems - 13th International Conference, OPODIS 2009, Proceedings
Pages	129-143
Number of pages	15
DOIs	https://doi.org/10.1007/978-3-642-10877-8_12
State	Published - 2009
Event	13th International Conference on Principles of Distributed Systems, OPODIS 2009 - Nimes, France Duration: Dec 15 2009 → Dec 18 2009

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	5923 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	13th International Conference on Principles of Distributed Systems, OPODIS 2009
Country/Territory	France
City	Nimes
Period	12/15/09 → 12/18/09

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-642-10877-8_12

Cite this

Bazzi, R., & Herlihy, M. (2009). Enhanced fault-tolerance through Byzantine failure detection. In Principles of Distributed Systems - 13th International Conference, OPODIS 2009, Proceedings (pp. 129-143). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5923 LNCS). https://doi.org/10.1007/978-3-642-10877-8_12

Enhanced fault-tolerance through Byzantine failure detection. / Bazzi, Rida; Herlihy, Maurice.
Principles of Distributed Systems - 13th International Conference, OPODIS 2009, Proceedings. 2009. p. 129-143 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5923 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Bazzi, R & Herlihy, M 2009, Enhanced fault-tolerance through Byzantine failure detection. in Principles of Distributed Systems - 13th International Conference, OPODIS 2009, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5923 LNCS, pp. 129-143, 13th International Conference on Principles of Distributed Systems, OPODIS 2009, Nimes, France, 12/15/09. https://doi.org/10.1007/978-3-642-10877-8_12

@inproceedings{3941cc71570e409181e757fc1e293605,

title = "Enhanced fault-tolerance through Byzantine failure detection",

abstract = "We consider a variant of the Byzantine failure model in which Byzantine processes are eventually detected and silenced, and investigate the fault-tolerance of the classical broadcast and agreement problems. We show that if all Byzantine processes are eventually detected, then it is possible to solve the broadcast problem in the presence of any number of Byzantine processes. If only a fraction of the Byzantine processes can be detected, then we show that it is possible to solve consensus (and broadcast) if the total number of processes is N ≥ 2f + 3F + 1, where f is the number of Byzantine processes that are eventually detected and F is the number of those that are never detected. We show that 2f + 3F + 1 is a lower bound to solve the consensus and broadcast problems.",

author = "Rida Bazzi and Maurice Herlihy",

year = "2009",

doi = "10.1007/978-3-642-10877-8_12",

language = "English (US)",

isbn = "3642108768",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "129--143",

booktitle = "Principles of Distributed Systems - 13th International Conference, OPODIS 2009, Proceedings",

note = "13th International Conference on Principles of Distributed Systems, OPODIS 2009 ; Conference date: 15-12-2009 Through 18-12-2009",

}

TY - GEN

T1 - Enhanced fault-tolerance through Byzantine failure detection

AU - Bazzi, Rida

AU - Herlihy, Maurice

PY - 2009

Y1 - 2009

N2 - We consider a variant of the Byzantine failure model in which Byzantine processes are eventually detected and silenced, and investigate the fault-tolerance of the classical broadcast and agreement problems. We show that if all Byzantine processes are eventually detected, then it is possible to solve the broadcast problem in the presence of any number of Byzantine processes. If only a fraction of the Byzantine processes can be detected, then we show that it is possible to solve consensus (and broadcast) if the total number of processes is N ≥ 2f + 3F + 1, where f is the number of Byzantine processes that are eventually detected and F is the number of those that are never detected. We show that 2f + 3F + 1 is a lower bound to solve the consensus and broadcast problems.

AB - We consider a variant of the Byzantine failure model in which Byzantine processes are eventually detected and silenced, and investigate the fault-tolerance of the classical broadcast and agreement problems. We show that if all Byzantine processes are eventually detected, then it is possible to solve the broadcast problem in the presence of any number of Byzantine processes. If only a fraction of the Byzantine processes can be detected, then we show that it is possible to solve consensus (and broadcast) if the total number of processes is N ≥ 2f + 3F + 1, where f is the number of Byzantine processes that are eventually detected and F is the number of those that are never detected. We show that 2f + 3F + 1 is a lower bound to solve the consensus and broadcast problems.

UR - http://www.scopus.com/inward/record.url?scp=72749110089&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=72749110089&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-10877-8_12

DO - 10.1007/978-3-642-10877-8_12

M3 - Conference contribution

AN - SCOPUS:72749110089

SN - 3642108768

SN - 9783642108761

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 129

EP - 143

BT - Principles of Distributed Systems - 13th International Conference, OPODIS 2009, Proceedings

T2 - 13th International Conference on Principles of Distributed Systems, OPODIS 2009

Y2 - 15 December 2009 through 18 December 2009

ER -

Enhanced fault-tolerance through Byzantine failure detection

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this