“Γ-accurate” failure detectors
The knowledge about failures needed to solve distributed agreement problems can be expressed in terms of completeness and accuracy properties of failure detectors introduced by Chandra and Toueg. The accuracy properties they have considered restrict the false suspicions that can be made by all the processes in the system. In this paper, we define “Γ-accurate” failure detectors, whose accuracy properties (only) restrict the false suspicions that can be made by a subset Γ of the processes. We discuss the relations between the classes of Γ-accurate failure detectors, and the classes of failure detectors defined by Chandra and Toueg. Then we point out the impact of these relations on the solvability of agreement problems.
Unable to display preview. Download preview PDF.
- 1.A. Basu, B. Charron-Bost and S. Toueg. Simulating Reliable Links with Unreliable Links in the Presence of Process Crashes. Proceedings of the 10th International Workshop on Distributed Algorithms, LNCS, Springer Verlag, October 1996.Google Scholar
- 2.O. Babaoglu, R. Davoli and A. Montresor. Failure Detectors, GroupMembership and View-Synchronous Communication in Partitionable Systems. Technical Report, University of Bologna, Computer Science Department. November 1995.Google Scholar
- 3.T. Chandra, V. Hadzilacos and S. Toueg. The weakest failure detector for solving consensus. Journal of the ACM, 43(4), July 1996. A preliminary version appeared in Proceedings of the 11th ACM Symposium on Principles of Distributed Computing, pp 147–159. ACM Press. August 1992.Google Scholar
- 4.T. Chandra and S. Toueg. Unreliable failure detectors for reliable distributed systems. Journal of the ACM, 34(1), pp 225–267, March 1996. A preliminary version appeared in Proceedings of the 10th ACM Symposium on Principles of Distributed Computing, pp 325–340. ACM Press. August 1991.CrossRefGoogle Scholar
- 5.R. Friedman, I. Keidar, D. Malki, K. Birman and D. Dolev. Deciding in Partitionable Networks, Technical Report, Cornell University, Computer Science Department. November 1995.Google Scholar
- 6.R. Guerraoui. Revisiting the relationship between Non Blocking Atomic Commitment and Consensus problems. Proceedings of the 9th International Workshop on Distributed Algorithms, pages 87–100, LNCS 972, Springer Verlag, September 1995.Google Scholar
- 7.R. Guerraoui and A. Schiper. Atomic Multicast harder than Atomic Broadcast. Technical Report, Ecole Polytechnique Fédérale de Lausanne, Computer Science Department. May 1996.Google Scholar
- 8.R. Guerraoui and A. Schiper. “Γ-Accurate” Failure Detectors. Technical Report, Ecole Polytechnique Fédérale de Lausanne, Computer Science Department. May 1996.Google Scholar
- 9.I. Keidar and D. Dolev. Increasing the Resilience of Atomic Commit, at No Additional Cost. Proceedings of the ACM Symposium on Principles of Database Systems, pages 245–254. ACM Press, May 1994.Google Scholar
- 10.L. Sabel and K. Marzullo. Election Vs. Consensus in Asynchronous Systems. Technical Report TR95-1488, Cornell Univ, 1995.Google Scholar
- 11.D. Skeen. NonBlocking Commit Protocols. Proceedings of the ACM SIGMOD International Conference on Management of Data, pages 133–142. ACM Press, 1981.Google Scholar
- 12.D. Skeen. A Quorum-Based Commit Protocol. Proceedings of the Berkeley Workshop on Distributed Data Management and Computer Networks, pages 69–80, Num 6, 1982.Google Scholar