Find White Papers
Home About Contact Help
Free Membership Member Login
Search the Library                  Advanced Search

The Definition of a Business Service for Customer Data Integration

IBM
By : IBM
INFORMATION
Published : Mar 01, 2006
Length : 20
Type : White Paper
 
Download Now
Save for Later
  Email This Page
Overview :

With the rapid growth in interest in customer data integration (CDI) applications and service oriented architectures (SOAs), many vendors have rushed to define their solutions as customer data integration (CDI), service-oriented applications. One of the key points of difference is each vendor’s definition of business service. This is a critical point for customers to understand, because the primary functionality and interface into a CDI application are its business services.

This paper will define different vendor approaches to developing business services and the relative cost implications of each approach for the customer.

View All Items By This Company
Browse Related Categories :

Analytical Applications

,

Best Practices

,

Business Intelligence

,

Business Intelligence

,

Customer Interaction Service

,

Data Integration

,

Data Management

,

Enterprise Applications

,

Information Management

,

Service Oriented Architecture

 
The key definition points for a business service are that it must:

- Represent a function

- Be offered at a granular, but functional, level

- Be offered at a larger-grain level as an aggregate of fine-grain services

- Permit integration through multiple technologies directly to the service (Web services) for both loose and tight coupling of the service, as opposed to forcing all integration through message queuing

- Be able to be combined with other services into a composite transaction

- Permit customization of business logic within the service

CDI business services represent customer data management functions. The primary responsibility of those functions is to maintain master customer data in the customer database. A service-oriented CDI application contains the following elements shown in the IBM WebSphere Customer Center example in Figure 1.

Large-grain business services are defined as customer data processes. Large-grain services comprise many fine-grain services and represent a significant function performed by the organization. Examples include adding a new customer or adding a contract and multiple new customers.

Fine-grain business services are defined as more atomic customer data processes or components of a larger unit of work. Fine-grain services encompass one or more objects (and therefore data tables in the database). Fine-grain services represent a level of functional abstraction above table-based create, read, update and delete functions. Examples include adding an address for a party or updating a party identifier.

An object model is defined as atomic functions. It is a specification of the objects intrinsic to the given application, including a description of the object characteristics (attributes), the object's functions, and a description of the static and dynamic relationships that exist between objects. At some point, these objects need to be persisted and, if they are persisted in a relational database, an object-relation mapping must occur. For data-centric applications (like CDI hubs), objects typically represent data functions and are a logical representation of the physical database for master customer data. Functions include creating, reading, updating and deleting data. Granular objects are not truly functional because they represent the database structure, not a process or functional usage of customer data. Objects are not business services.

A database is defined as the database in which customer master data is stored in multiple tables.


Service-oriented application approach

Overview

Service-oriented CDI applications' central design point is the business service that will be used by multiple business applications. These applications were designed from the onset as service-oriented applications. All functionality was designed as a callable service. The result is that different functionality is made available when the original design point is building services as compared to service-enabling of an existing application.

The definition of a business service for customer data integration

Key points

Service-oriented CDI applications have important differences from either application-suite or tool-based approaches, including:

- The middle-ground definition of a service: SOA CDI hubs deliver both large- and fine-grain services. Those services offer varying levels of functionality that represent microflows, which are business functions that are more atomic than a business process that would be modeled in an enterprise application integration (EAI) workflow or a business process modeling (BPM) tool. Microflows are larger than granular data-table-level data functions (adding and updating data within a database table). In summary, microflows are atomic to a functional level for customer data management.

- Accessibility: SOA CDI hubs offer multiple access methods to directly call the business services. These include Web services, XML, EAI publish-subscribe interfaces and object-level interfaces. The key difference between SOA CDI hubs and other approaches is the fact that the services can be directly called by an application or integration hub-previously accessible only through an unnecessary, middleman EAI or work-flow tool. This offers greater flexibility in deployment and integration of the CDI hub shown in Figure 3.
Search the Library                  Advanced Search
About Us Contact Us List Your Papers Partner With Us Site Map