A new primary lazy update propagation protocol for replicated databases

This paper proposes a new primary lazy update protocol, PTCS (Primary Transaction Commit Schedule). In the PTCS protocol, a serializable primary transaction schedule is generated firstly and then the secondary transactions are committed according to the serializable primary transaction schedule. PTCS protocol can guarantee serializability if the data copy graph contains no directed circles. It can also be extended to eliminate all requirements on the data copy graph. Compared to earlier works, PTCS protocol not only imposes a much weaker requirement on the data placement, but also avoids the deadlock caused by transaction waits and extra message overhead. The performance experiments show that the degradation of the performance caused by the replica management of the PTCS protocol is tolerable.

[1]  S. S. Ravi,et al.  Deferred updates and data placement in distributed databases , 1996, Proceedings of the Twelfth International Conference on Data Engineering.

[2]  Dennis Shasha,et al.  The dangers of replication and a solution , 1996, SIGMOD '96.

[3]  Avishai Wool,et al.  Replication, consistency, and practicality: are these mutually exclusive? , 1998, SIGMOD '98.

[4]  Gong Yunzhan,et al.  Minimizing the copy graph to improve scalability , 2003, Proceedings of the Fourth International Conference on Parallel and Distributed Computing, Applications and Technologies.

[5]  Rajeev Rastogi,et al.  Update propagation protocols for replicated databates , 1999, SIGMOD '99.

[6]  Henry F. Korth,et al.  Replication and consistency: being lazy helps sometimes , 1997, PODS.