Slice extends OpenJPA for distributed databases. Slice plugs-in to OpenJPA runtime as a single library and can be activated by configuring a persistence unit for multiple databases. Once configured for Slice, an existing OpenJPA application can transact against multiple databases in the same transaction. The application queries will be executed in parallel against all the databases and any update will be committed to the appropriate databases.

Java Persistence API (JPA) is a industry-standard specification for persistence of Java objects to relational database. JPA standardizes the complex object-relational mapping problem and is supported by mature implementations from vendors such as BEA Kodo, JBoss Hibernate or Oracle Toplink. JPA is also the persistence service for EJB 3.0 in JEE environment. OpenJPA - a top level Apache Open Source Project - is a feature-rich, production-quality implementation of JPA. OpenJPA supports numerous extended features over JPA standards and its well-crafted architecture allows a sophisticated plug-in framework for addition of new features. Slice is one such plug-in extension to enable any OpenJPA-based application to work with distributed, horizontally-partitioned, possibly heterogeneous databases without any change to application code or the persistent domain model.

Enterprise applications are increasingly using a distributed databases. The reasons for distributed, often horizontally-partitioned databases can be to counter massive data growth, or to support multiple external clients on a hosted platform and in many other scenarios that may benefit from data partitioning. Any JPA-based user application has to address serious technical and conceptual challenges if it tries to directly interact with a set of physical databases within a single transaction. Slice encapsulates this complexity of distributed database interaction via the abstraction of a virtual database which internally manages multiple physical databases henceforth referred as database slice. This virtualization of distributed databases approach makes OpenJPA's object management kernel and the user application to work in the same way as in case of a single physical database. Via this database virtualization approach, Slice effectively provides the user application an object-oriented view over a distributed set of, possibly heterogeneous, databases,

This document describes how to use Slice to enable OpenJPA applications for distributed database environment. This document also discusses the architecture and critical design aspects of Slice.

This document is intended for developers interested in building distributed database applications using OpenJPA. This document assumes familiarity with OpenJPA usage and distributed database environment. This document is also targeted for developers who are interested in extending OpenJPA or its basic architectural aspects.

&homer;