The K2 Persistent State Service (PSS) is a fully compliant implementation of OMG's specification for Persistent State Service. It presents persistent information as storage objects stored in storage homes which themselves are stored in datastores, an entity that manages data, for example a database, a set of files, a schema in a relational database.

The Persistent State Service provides a language for describing the states to persist in a datastore. This language is called as Persistent State Definition Language(PSDL). PSDL is a superset of OMG IDL, and is a declarative language like IDL.

The K2 Persistent State Service provides two ways to define the datastore schema and the
application programming interface of storage object instances in a datastore with this
schema:

• Using the Persistent State Definition Language (PSDL)
  
  
• Transparent Persistence using C++.
  

The state information to persist is defined using PSDL language. The K2 Persistent State Service provides a PSDL compiler for programmers to generate persistent code for state information. The K2 PSDL compiler generates database independent code for various database like Berkley DB , Oracle etc.

The fundamental concept of Persistent State Service is defined using few constructs like abstract storagetype , abstract storagehome, storagetype, storagehome and catalog. A storagetype definition is like a storage object instance ( like a programming language ( c++, Java) class instances. This storage object instance is bound to a single record in a database table and provides direct access to the state

information. Similarly, storagehome definition is like a language class instance which is provided by the catalogs. To access a storage object, you need a logical connection between your process and Datastore that contains the storagehome of this storage object type.

Features of K2 Persistent State Service

1. Support for Oracle connector.
   An application can persist its data in Oracle Database, through Oracle connector.
   
   
2. Support for BerkleyDB Connector
   An application can persist its data in BerkleyDB Database, through BerkleyDB connector
   
   
3. Support for embedded storage objects.
   The relationship between different tables in a RDBMS which doesn't have a primary key, can be specified by embedding one storage objects inside another.
   
   
4. Support for reference to another storage object.
   The relationship between different tables in a RDBMS can be specified by storing a reference of one storage object inside another storage object through the ref keyword.
   
   
5. InMemory Caching mechanism
   The persisted data is cached in memory, for higher efficiency and improve data access performance.
   
   
6. Delayed writing to the database
   Several storage object can be modified and flushed to the database in single call, This reduces the access frequency to the datastores.
   
   
7. Logging mechanism
   Application specific messages can be logged in a file or to the console as desired.
   
   
8. PSDL compiler
   The K2PSDL compiler generates database independent persistence code. So the developer need not be concerned of specific database vendors while developing the application.