A Probabilistic Analysis of Snapshot Isolation with Partial Replication

Snapshot isolation has received a considerable amount of attention in the context of full database replication. Such popularity is mainly because read-only transactions executing under snapshot isolation are never blocked or aborted. In partial replication, where each replica holds only a part of the database, transactions may require access to remote databases. Each remote read operation of the transaction must execute in a consistent global database snapshot as the local operations; if such a snapshot is not available, the transaction must be aborted. In this paper we are interested in the effects of distributed transactions on the abort rate of partially replicated snapshot isolation systems. We present a simple probabilistic analysis of transaction abort rates for two different concurrency control mechanisms: lock- and version-based. The former models the behavior of a replication protocol providing one-copy-serializability; the latter models snapshot isolation. Our analysis reveals that in the version-based system the execution abort rate decreases exponentially as the number of data versions available increases. As a consequence, in all cases considered, two versions of each data item were sufficient to eliminate aborts due to distributed transactions.

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

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

[3]  Ricardo Jiménez-Peris,et al.  Boosting Database Replication Scalability through Partial Replication and 1-Copy-Snapshot-Isolation , 2007, 13th Pacific Rim International Symposium on Dependable Computing (PRDC 2007).

[4]  U. Fritzke,et al.  Transactions on partially replicated data based on reliable and atomic multicasts , 2001, Proceedings 21st International Conference on Distributed Computing Systems.

[5]  Rachid Guerraoui,et al.  Exploiting Atomic Broadcast in Replicated Databases , 1998, Euro-Par.

[6]  Patrick Valduriez,et al.  Consistency management for partial replication in a high performance database cluster , 2005, 11th International Conference on Parallel and Distributed Systems (ICPADS'05).

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

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

[9]  Fernando Pedone,et al.  Sprint: a middleware for high-performance transaction processing , 2007, EuroSys '07.

[10]  Bettina Kemme,et al.  Database replication for clusters of workstations , 2000 .

[11]  Fernando Pedone,et al.  Partial replication in the Database State Machine , 2001, Proceedings IEEE International Symposium on Network Computing and Applications. NCA 2001.

[12]  Francesc D. Muñoz-Escoí,et al.  SIPRe: a partial database replication protocol with SI replicas , 2008, SAC '08.

[13]  Fernando Pedone.  The database state machine and group communication issues , 1999 .

[14]  Fernando Pedone,et al.  Optimistic Algorithms for Partial Database Replication , 2006, OPODIS.

[15]  Sameh Elnikety,et al.  Tashkent+: memory-aware load balancing and update filtering in replicated databases , 2007, EuroSys '07.

[16]  Divyakant Agrawal,et al.  Partial database replication using epidemic communication , 2002, Proceedings 22nd International Conference on Distributed Computing Systems.