Abstract
The Distributed Consensus problem involves n processors each of which holds an initial binary value. At most t processors may be faulty and ignore any protocol (even behaving maliciously), yet it is required that the non-faulty processors eventually agree on a value that was initially held by one of them. This paper presents consensus protocols that tolerate arbitrary faults, are early-stopping (i.e., run for a number of rounds proportional to the number of faults f that actually occur during their execution), and are optimal in various measures.
Our first contribution is an early-stopping consensus protocol that is simultaneously optimal in round complexity (i.e., min(t+1, f+2) rounds) and number of processors (i.e., n > 3t). This settles a long-standing open question [DRS, 1982]. These bounds were not known to be attainable even with the use of authentication. Since consensus is not attainable with n <- 3t we provide a definitive answer to the problem of early-stopping consensus. Previous protocols for this problem that achieved round optimality required n=Ω(t 2) [C, DRS], n > 6t [MW] and n > 4t [BGP]. Instrumental in achieving this result is the new safe message reconstruction technique, which we expect to be of broader applicability. The previous protocol is round- and processor-optimal, but not efficient. Our second contribution is a pair of optimal early-stopping consensus protocols that use messages of constant size. No previously existing early-stopping protocol, whether exactly or only asymptotically optimal in some measures, has used messages of constant size. The only other existing early-stopping protocol with constant-sized messages requires n=Ω(t 2) [C].
Finally, we indicate how to extend one of the previous protocols to be optimal in total bit (and message) complexity and number of processors. This is the first protocol that is both optimal in these measures and early-stopping.
This is a preview of subscription content, log in via an institution to check access.
Preview
Unable to display preview. Download preview PDF.
References
A. Bar-Noy, D. Dolev, C. Dwork and H.R. Strong, “Shifting gears: changing algorithms on the fly to expedite Byzantine Agreement,” Proc. 6th PODC, pp. 42–51, August 1987.
P. Berman and J.A. Garay, “Asymptotically Optimal Distributed Consensus,” Proc. ICALP 89, LNCS Vol. 372, pp. 80–94, July 1989.
P. Berman and J.A. Garay, “Distributed Consensus with n = (3 + ε)t Processors,” Proc. 5th International Workshop on Dist. Algorithms, LNCS, Springer-Verlag, October 1991.
P. Berman, J.A. Garay and K.J. Perry, “Towards Optimal Distributed Consensus,” Proc. 30th FOCS, pp. 410–415, October/November 1989.
B. Coan, “Efficient agreement using fault diagnosis,” Proc. 26th Allerton Conf. on Comm., Control and Computing, pp. 663–672, 1988.
D. Dolev and R. Reischuk, “Bounds of Information Exchange for Byzantine Agreement,” JACM, Vol. 32, No. 1, pp. 191–204, 1985.
D. Dolev, R. Reischuk and H.R. Strong, “Eventual is Earlier than Immediate,” in Proc. 23rd STOC, 1982. Revised version appears in “Early Stopping in Byzantine Agreement,” JACM, Vol. 37, No. 4 (1990), pp. 720–741.
D. Dolev and H.R. Strong, “Authenticated Algorithms for Byzantine Agreement,” in SIAM Journal of Computing, Vol. 12, pp. 656–666, 1983.
P. Feldman and S. Micali, “Optimal Algorithms for Byzantine Agreement”, Proc. 20th STOC, pp. 148–161, May 1988.
V. Hadzilacos and J. Halpern, “Message-Optimal Protocols for Byzantine Agreement,” Proc. 10th PODC, pp. 309–324, August 1991.
J. Halpern and Y. Moses, “Knowledge and common knowledge in a distributed environment,” JACM, Vol. 37, No. 3 (1990), pp. 549–587.
L. Lamport and M. Fischer, Byzantine Generals and Transaction Commit Protocols, Opus 62, SRI International, April 1982.
L. Lamport, R.E. Shostak and M. Pease, “The Byzantine Generals Problem,” ACM ToPLaS, Vol. 4, No. 3, pp. 382–401, July 1982.
Y. Moses and O. Waarts, “Coordinated Traversal: (t+1)-Round Byzantine Agreement in Polynomial Time,” Proc. 29th FOCS, pp. 246–255, October 1988.
K.J. Perry and S. Toueg, “Distributed agreement in the presence of processor and communication faults,” IEEE Trans. on Software Engineering, Vol. 12, No. 3, pp. 477–482, March 1986.
S. Toueg, K.J. Perry and T.K. Srikanth, “Fast Distributed Agreement,” SLAM Journal of Computing, Vol. 16, No. 3, pp. 445–457, June 1987.
O. Waarts, “Coordinated Traversal: Byzantine Agreement in polynomial time,” M.Sc. Thesis, Weizmann Institute of Science, Rehovot, Israel, August 1988.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1992 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Berman, P., Garay, J.A., Perry, K.J. (1992). Optimal early stopping in distributed consensus. In: Segall, A., Zaks, S. (eds) Distributed Algorithms. WDAG 1992. Lecture Notes in Computer Science, vol 647. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-56188-9_15
Download citation
DOI: https://doi.org/10.1007/3-540-56188-9_15
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-56188-0
Online ISBN: 978-3-540-47484-5
eBook Packages: Springer Book Archive