IBM BPM, V8.0.1, All platforms > Administer applications and processes in the runtime environment > Manage relationships
The relationship service
The relationship service maintains relationships and roles in the system. It manages relationship and role definitions and metadata and makes it possible to specify the definition of a relationship and manipulate the instances derived from the definition.
The relationship service makes it possible to capture relationships across different objects. Participants in the relationship are distinguished by the roles they serve.
For example, a Person object "Joe" can have an ownership relationship with a Car object "Subaru with license plate XYZ 123."In this example, Joe participates in the relationship with the role "owner"while the car participates in the relationship under the role "owned object".
Relationship and role definitions
Relationships and roles are described in definitions that you design through the graphical interface of the relationship editor tool in IBM Integration Designer. The relationship definition is a template that describes what the relationship should look like, identifying the roles each participant in the relationship can assume. The role definition captures the structure and constraint requirements for the participants. Relationship definitions are stored as XML files that are deployed as part of a Java EE application to a particular server.
For detailed background and task information on creating relationships, identifying relationship types, and using the relationship editor, see the 
IBM Integration Designer Information Center.
How relationships work
At run time, when maps or other IBM BPM components run and need a relationship instance, the instances of the relationships are either created or retrieved, depending on the scenario. The relationship and role instance data can be manipulated through three means:
- IBM BPM component Java™ snippet invocations of the relationship service APIs
- Relationship transformations in the IBM BPM business object mapping service
- Using the relationship manager tool
The relationship and role instance data is saved in relationship tables that are stored in the database in the default data source that you specify when you configure the relationship service.
The relationship service runs on each server at the cell level. The Relationship Manager home page About section shows the number of servers in the cell that are running relationship services; the Relationships section shows each server name that is running relationship services. Before working with relationship instances, you need to select the server that has the instances of the relationships and roles you want to manage.
Relationship definitions that are contained in user-defined shared libraries will be created in the server at startup time, even if those shared libraries are not associated with any server or application. You can use the relationship manager to manage these relationships.
Tip: To view all of the relationships managed by the relationship service, use the Relationship Manager. Click Integration Applications > Relationship Manager > Relationship Services configuration > Relationships.