Efficient management of replicated data
- First Online:
We propose in this paper a protocol aimed at managing replicated data with many copies more efficiently. Our protocol, called Majority Consensus Voting with Tokens generalizes the primary copy approach by allowing for more than one primary copy. Since primary copies are identified by the ownership of a token, the protocol allows for easy transfer of tokens from one copy to another to respond to changing access patterns. Tokens owned by failed primary copies can also be reassigned to some live copy to enhance data availability.
Upper and lower bounds for the availability of replicated data objects managed by the MCVT protocol are derived under very general probabilistic hypotheses. We show that MCVT significantly improves upon the availability and reliability offered by the primary copy protocol without incurring the high overhead of most other extant protocols.
Unable to display preview. Download preview PDF.
- [BGS86]D. Barbara, H. Garcia-Molina and A. Spauster, “Policies for Dynamic Vote Reassignment,” Proc. Sixth International Conference on Distributed Computing Systems (1986), pp. 37–44.Google Scholar
- [BaHe84]Barslow, R. E. and K. D. Heidtmann, Computing k-out-of-n System Reliability. IEEE Trans. on Reliability, R-33, 4 (Oct. 1984), 322–323.Google Scholar
- [BeGo84]P. A. Bernstein and N. Goodman, “An Algorithm for Concurrency Control and Recovery in Replicated Distributed Databases,” ACM Trans. on Database Systems, Vol. 9, No. 4 (Dec. 1984), 596–615.Google Scholar
- [CLP87]J. L. Carroll, D. D. E. Long and J.-F. Pâris, “Block-Level Consistency of Replicated Files.” Proc. Seventh International Conference on Distributed Computing Systems (Sept. 1987), pp. 146–153.Google Scholar
- [DaBu85]D. Davcev and W.A. Burkhard, “Consistency and Recovery Control for Replicated Files,” Proc. Tenth ACM Symposium on Operating System Principles, (1985), pp. 87–96.Google Scholar
- [Elli77]C. A. Ellis, “Consistency and Correctness of Duplicate Database Systems,” Operating Systems Review, 11, 1977.Google Scholar
- [Giff79]D. K. Gifford, “Weighted Voting for Replicated Data,” Proc. Seventh ACM Symposium on Operating System Principles, (1979), pp. 150–161.Google Scholar
- [Good83]N. Goodman, D. Skeen, A. Chan, U. Dayal, R. Fox and D. Ries, “A Recovery Algorithm for a Distributed Database System,” Proc. Second ACM Symposium on Principles of Database Systems, (1983), pp. 8–15.Google Scholar
- [Jajo87]S. Jajodia, “Managing Replicated Files in Partitioned Distributed Database Systems,” Proc. Third International Conference on Data Engineering, (Feb. 1987).Google Scholar
- [LoPa87]D.D.E. Long and J.-F. Pâris, “On Improving the Availability of Replicated Files,” Proc. Sixth Symposium on Reliability in Distributed Systems and Database Systems, (1987), pp. 77–83.Google Scholar
- [LoPa88]D. D. E. Long and J.-F. Pâris, “Regeneration Protocols for Replicated Files,” submitted for publication.Google Scholar
- [NoAn87]J. D. Noe and A. Andreassian, “Effectiveness of Replication in Distributed Computing Networks,” Proc. Seventh International Conference on Distributed Computing Systems, (1987), pp. 508–513.Google Scholar
- [PaLo88]J.-F. Pâris and D. D. E. Long, “Efficient Dynamic Voting Algorithms,” Proc. Fourth International Conference on Data Engineering, Los Angeles, Calif. (February 1988), pp. 268–275.Google Scholar
- [Pari87]J.-F. Pâris, “Operational Analysis of Redundant Systems,” Proc. International Workshop on Modelling Techniques and Performance Evaluation (March 1987), pp. 41–48.Google Scholar
- [PNP86]C. Pu, J. D. Noe and A. Proudfoot, “Regeneration of Replicated Objects: A Technique and its Eden Implementation,” Proc. Second International Conference on Data Engineering, Los Angeles, Calif. (February 1986), pp. 175–187.Google Scholar
- [Ston79]M. Stonebraker, “Concurrency Control and Consistency of Multiple Copies of Data in Distributed INGRES,” IEEE Transactions on Software Engineering, SE-5, 3 (May 1979), 188–194.Google Scholar