Danny Dolev
Abstract
Reaching agreement is a primitive of distributed computing. Whereas this poses no problem in an ideal, failure-free environment, it imposes certain constraints on the capabilities of an actual system: A system is viable only if it permits the existence of consensus protocols tolerant to some number of failures. Fischer et al. have shown that in a completely asynchronous model, even one failure cannot be tolerated. In this paper their work is extended: Several critical system parameters, including various synchrony conditions, are identified and how varying these affects the number of faults that can be tolerated is examined. The proofs expose general heuristic principles that explain why consensus is possible in certain models but not possible in others.
- 1 AGHILI, H., ASTRAHAN, M., FINKELSTEIN, S., KIM, W., MCPHERSON, J., SCHKOLNICK, M., AND STRONG, R. A prototype for a highly available database system. Rep. RJ 3755, IBM Research Division, San Jose, Calif., 1983.Google Scholar
- 2 ATTIYA, C., DOLEV, D., AND GIL, J. Asynchronous Byzantine consensus. In Proceedings of the 3rd Annual ACM Symposium on Principles of Distributed Computing (Vancouver, B.C., Canada, Aug. 27-29). ACM, New York, 1984, pp. 119-133. Google Scholar
- 3 BEN-OR, M. Another advantage of free choice: Completely asynchronous agreement protocols. In Proceedings of the 2nd Annual ACM Symposium on Principles of Distributed Computing (Montreal, Quebec, Canada, Aug. 17-19). ACM, New York, 1983, pp. 27-30. Google Scholar
- 4 BEN-OR, M. Fast asynchronous Byzantine agreement. In Proceedings of the 4th Annual ACM Symposium on Principles of Distributed Computing (Minaki, Ontario, Canada, Aug. 5-7). ACM, New York, 1985, pp. 149-151. Google Scholar
- 5 BRACHA, G. An asynchronous t(n - 1)/3.I-resilient consensus protocol. In Proceedings of the 3rd Annual ACM Symposium on Principles of Distributed Computing (Vancouver, B.C., Canada, Aug. 27-29). ACM, New York, 1984, pp. 154-162. Google Scholar
- 6 DOLEV, D., AND REmCHUK, R. Bounds on information exchange for Byzantine agreement. In Proceedings of the ACM SIGACT-SIGOPS Symposium on Principles of Distributed Computing (Ottawa, Canada, Aug. 18-20). ACM, New York, 1982, pp. 132-140. Google Scholar
- 7 DOLEV, D., AND STRONG, H. R. Authenticated algorithms for Byzantine agreement. SIAM J. Comput. 12 (1983), 656-666.Google Scholar
- 8 DOLEV, D., REISCHUK, R., AND STRONG, H.R. Eventual is earlier than immediate. In Proceedings of the 23rd Annual IEEE Symposium on Foundations of Computer Science (Chicago, I11., Nov. 3-5). IEEE, New York, 1982, pp. 196-203.Google Scholar
- 9 DWORK, C., LYNCH, L., AND STOCKMEYER, L. Consensus in the presence of partial synchrony. IBM Res. Rep. RJ 4892, IBM Research Division, San Jose, Calif., Oct. 1985.Google Scholar
- 10 FISCHER, M. J., LYNCH, N. A., AND PATERSON, M.S. Impossibility of distributed consensus with one faulty process. J. ACM 32 (1985), 374-382. Google Scholar
- 11 LAMPORT, L., SHOSTAK, R., AND PEASE, M. The Byzantine generals problem. ACM Trans. Program. Lang. Syst. 4, 3 (July 1982), 382-401. Google Scholar
- 12 PEASE, M., SHOSTAK, R., AND LAMPORT, L. Reaching agreement in the presence of faults, j. ACM 27, 2 (Apr. 1980), 228-234. Google Scholar
- 13 RAmN, M.O. Randomized Byzantine generals. In Proceedings of the 24th Annual IEEE Symposium on Foundations of Computer Science (Tucson, Ariz., Nov. 7-9). IEEE, New York, pp. 403-409. Google Scholar
- 14 TOUEG, S. Randomized Byzantine agreements. In Proceedings of the 3rdAnnualACMSymposium on Principles of Distributed Computing (Vancouver, B.C., Canada, Aug. 27-29). ACM, New York, 1984, pp. 163-178. Google Scholar
Index Terms
- On the minimal synchronism needed for distributed consensus
Recommendations
On the minimal synchronism needed for distributed consensus
SFCS '83: Proceedings of the 24th Annual Symposium on Foundations of Computer ScienceReaching agreement is a primitive of distributed computing. While this poses no problem in an ideal, failure-free environment, it imposes certain constraints on the capabilities of an actual system: a system is viable only if it permits the existence of ...
Minimal Synchrony for Byzantine Consensus
PODC '15: Proceedings of the 2015 ACM Symposium on Principles of Distributed ComputingSolving the consensus problem requires in one way or another that the underlying system satisfies some synchrony assumption. Considering an asynchronous message-passing system of n processes where (a) up to t< n/3 may commit Byzantine failures, and (b) ...
Resilient distributed consensus
MobiHoc '12: Proceedings of the thirteenth ACM international symposium on Mobile Ad Hoc Networking and ComputingConsensus algorithms allow a set of nodes to reach an agreement on a quantity of interest. For instance, a consensus algorithm may be used to allow a network of sensors to determine the average value of samples collected by the different sensors. ...
Reviews
Greg Speegle
Reaching agreement in the presence of failures by a distributed system is an interesting problem. This paper defines the conditions that are sufficient for agreement to occur and some conditions that make agreement impossible. The paper identifies five important components of a distributed system and constructs a favorable and an unfavorable condition for each component. Proofs are presented that show the resiliency possible under the 32 possible combinations of conditions of the components. The paper is very formal and contains many proofs. These proofs are presented in an organized style, but the notation required to present them is difficult to read. The primary value of this work is the clear definitions of what is and is not possible in reaching agreement. Of course, as in any case analysis, some systems may not fit the model used perfectly; but with the techniques presented and the broad range of situations covered, one could derive similar results for other works with relatively easy proofs.
Access critical reviews of Computing literature here
Become a reviewer for Computing Reviews.
Comments