Introduction to Slice

+-------------------------------------------------------------------------
 Slice extends {{{http://openjpa.apache.org}OpenJPA}} to work with horizontally-partitioned, distributed databases.
+-------------------------------------------------------------------------

[images/slice_small]        

 Slice abstracts horizontally-partitioned, distributed databases as a single 
 virtual database for OpenJPA so that the user application can use all OpenJPA 
 features without the technical challenges of handling multiple databases.
 {{{http://en.wikipedia.org/wiki/Partition_(database)}Horizontal partitioning}}
 refers to partitioning of the rows of a table across many databases. 
 We refer these underlying databases as <Slice> in this context.
 
 The salient features of Slice are: 
   
   [[1.]] configured as a plug-in to OpenJPA runtime
   
   [[2.]] no change to the application or to the persistent domain model 

   [[3.]] application decides the distribution of newly persistent instances
   across multiple databases
   
   [[4.]] each slice can be configured independently, thereby opening up the
   possibility where data is distributed across heterogeneous databases. 
   
 The <current> limitations are
 
   [[1.]] no relationships across databases; hence the closure of 
   an object graph must be collocated in the same database
   
   [[2.]] no two-phase commit protocol across multiple databases when local
   transaction is local. If OpenJPA is configured for a JTA transaction, then 
   the underlying slices participate in global transaction.

   [[3.]] no sorting of query result across multiple databases.