Correctness Criteria for Database Replication: Theoretical and Practical Aspects

In this paper we investigate correctness criteria for replicated databases from the client's perspective and present their uniform characterization. We further study the effects of different consistency degrees in the context of three middleware-based replication protocols: primary-backup, optimistic update-everywhere and BaseCON , a simple yet fault-tolerant middleware-based replication protocol that takes advantage of workload characterization techniques to increase the parallelism in the system. We present three variants of BaseCON, one for each correctness criterion discussed, and analyze its behavior in case of failures and false suspicions. We have implemented the correctness properties in all three protocols considered and show experimentally that stronger consistency does not necessarily imply worse performance. On the contrary, two of the three replication protocols evaluated show no significant performance divergence under the chosen benchmark while ensuring different consistency criterion.

[1]  Luis Irún-Briz,et al.  Managing Transaction Conflicts in Middleware-based Database Replication Architectures , 2006, 2006 25th IEEE Symposium on Reliable Distributed Systems (SRDS'06).

[2]  Rachid Guerraoui,et al.  The Database State Machine Approach , 2003, Distributed and Parallel Databases.

[3]  GrayJim,et al.  A critique of ANSI SQL isolation levels , 1995 .

[4]  ShashaDennis,et al.  The dangers of replication and a solution , 1996 .

[5]  Francisco Moura,et al.  Group-Based Replication of On-Line Transaction Processing Servers , 2005, LADC.

[6]  Hector Garcia-Molina,et al.  Overview of multidatabase transaction management , 1992, The VLDB Journal.

[7]  Ricardo Jiménez-Peris,et al.  Middleware based data replication providing snapshot isolation , 2005, SIGMOD '05.

[8]  Jim Gray,et al.  A critique of ANSI SQL isolation levels , 1995, SIGMOD '95.

[9]  Luís E. T. Rodrigues,et al.  The GlobData Fault-Tolerant Replicated Distributed Object Database , 2002, EurAsia-ICT.

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

[11]  Gustavo Alonso,et al.  MIDDLE-R: Consistent database replication at the middleware level , 2005, TOCS.

[12]  Fernando Pedone,et al.  Pronto: a fast failover protocol for off-the-shelf commercial databases , 2000, Proceedings 19th IEEE Symposium on Reliable Distributed Systems SRDS-2000.

[13]  Kenneth Salem,et al.  Lazy database replication with ordering guarantees , 2004, Proceedings. 20th International Conference on Data Engineering.

[14]  Gustavo Alonso,et al.  Ganymed: Scalable Replication for Transactional Web Applications , 2004, Middleware.

[15]  Sam Toueg,et al.  Unreliable failure detectors for reliable distributed systems , 1996, JACM.

[16]  Willy Zwaenepoel,et al.  C-JDBC: Flexible Database Clustering Middleware , 2004, USENIX Annual Technical Conference, FREENIX Track.

[17]  Marta Patiño-Martínez.  Consistent Database Replication at the Middleware Level , 2005 .

[18]  A Min Tjoa,et al.  EurAsia-ICT 2002: Information and Communication Technology , 2002, Lecture Notes in Computer Science.

[19]  Alan L. Cox,et al.  Conflict-Aware Scheduling for Dynamic Content Applications , 2003, USENIX Symposium on Internet Technologies and Systems.

[20]  Fernando Pedone,et al.  Database replication using generalized snapshot isolation , 2005, 24th IEEE Symposium on Reliable Distributed Systems (SRDS'05).

[21]  Chun Zhang,et al.  Storing and querying ordered XML using a relational database system , 2002, SIGMOD '02.

[22]  José Pereira,et al.  Revisiting 1-copy equivalence in clustered databases , 2006, SAC '06.