Symmetric Tree Replication Protocol for Efficient Distributed Storage System

In large distributed systems, replication of data and service is needed to decrease the communication cost, increase the overall availability, avoid single server bottleneck, and increase the reliability. Tree quorum protocol is one of the replication protocols allowing very low read cost in the best case but has some drawbacks such that the number of replicas grows rapidly as the level increases and root replica becomes a bottleneck. In this paper we propose a new replication protocol called symmetric tree protocol which efficiently solves the problems. The proposed symmetric tree protocol also requires much smaller read cost than the previous protocols. We conduct cost and availability analysis of the protocols, and the proposed protocol displays comparable read availability to the tree protocol using much smaller number of nodes. It is thus effective to be applied to survival storage system.

[1]  Robert H. Thomas,et al.  A Majority consensus approach to concurrency control for multiple copy databases , 1979, ACM Trans. Database Syst..

[2]  Alan L. Cox,et al.  Data replication strategies for fault tolerance and availability on commodity clusters , 2000, Proceeding International Conference on Dependable Systems and Networks. DSN 2000.

[3]  David K. Gifford,et al.  Weighted voting for replicated data , 1979, SOSP '79.

[4]  Divyakant Agrawal,et al.  The Tree Quorum Protocol: An Efficient Approach for Managing Replicated Data , 1990, VLDB.

[5]  Danco Davcev,et al.  A Dynamic Voting Scheme in Distributed Systems , 1989, IEEE Trans. Software Eng..

[6]  Satish K. Tripathi,et al.  An Analysis of the Average Message Overhead in Replica Control Protocols , 1996, IEEE Trans. Parallel Distributed Syst..

[7]  Makoto Takizawa,et al.  Group protocol for quorum-based replication , 2000, Proceedings Seventh International Conference on Parallel and Distributed Systems (Cat. No.PR00568).

[8]  Mostafa H. Ammar,et al.  The grid protocol: a high performance scheme for maintaining replicated data , 1990, [1990] Proceedings. Sixth International Conference on Data Engineering.

[9]  Bernd Freisleben,et al.  Designing multi-level quorum schemes for highly replicated data , 1991, [1991] Proceedings Pacific Rim International Symposium on Fault Tolerant Systems.

[10]  Philip A. Bernstein,et al.  An algorithm for concurrency control and recovery in replicated distributed databases , 1984, TODS.