LESSON - Enterprise Information Integration: A Technology for Providing Integrated Views
By Ashok Swaminathan, Director, Product Management—Sybase Data Integration, Sybase, Inc.
For organizations to extract value from multiple data sources quickly and efficiently, they must be able to integrate current data from heterogeneous sources. Three methods for developing this capability include:
- Accessing the data sources directly. However, keep the following in mind:
- Application developers must know what data to access and where and how to access it. Writing access logic (versus business logic) takes time. And making changes in data sources (adding or modifying tables or columns) means that the consuming applications must also be modified.
- Applications are tightly coupled with application servers and databases. Data access logic is therefore typically buried in the application logic, leaving little scope for reusability.
- Data owners have no easy way to see what parts of their data are used by applications, and no convenient way to control those applications’ access to that data.
- Using ETL technology to populate a data warehouse. Extract data from the operation store (usually through batch processing), transform, and make the data suitable for consumption by the decision support processes. The target is a data warehouse or operational data store. This approach has advantages, but is not optimal for current access to heterogeneous sources of data.
- Using enterprise application integration (EAI) technology to integrate the business processes of two or more applications via their APIs. EAI provides process-to-process integration and helps execute transactions and process flows. However, it’s not meant to process complex queries across data sources, aggregate the results, and serve them up as a result set.
While these technologies each serve different purposes, EII is best when users want to create unified views of data without making copies of the source information.
Enterprise Information Integration (EII)
EII solutions provide a means for integrating data in real time, allowing standardized access to the integrated data through a single, federated data layer. EII technology is ideal for applications that require current data that is selectively integrated from many sources:
- There is no need to rewrite applications to integrate a new data source or if the physical location of the data sources changes. New sources can be incrementally added to the federated data layer without affecting existing applications.
- Developers can find queries or services that have already been developed, and if appropriate, reuse them rather than recreate them. EII provides significant savings because developers can focus on the business logic rather than the access code.
- Data owners can see what aspects of their data they’ve made available to consuming applications. They can control that availability in terms of access authorization and the load on their data sources. Plus, they can track how this data is used.
Typical Uses of EII Technology
EII’s distinguishing characteristic is its ability to access and integrate current data from heterogeneous sources. How? A virtualized data federation layer delivers data to applications in real time from original sources. Various vertical segments increasingly use EII for business decisions:
- Banking. Banks provide multiple lines of service: checking, home equity loans, brokerage services, etc. Banks want a complete view of customer activities across lines of business from multiple divisions and applications. Most banking data is in silos, making EII a natural choice for integration.
- Securities Trading. As hedge funds trade across asset classes, traders need a complete view that incorporates up-to-the-minute data. Typically, different applications generate trading data, and different databases store it. EII allows integration from different sources, providing complete trading information across asset classes.
- Single Customer View. As organizations roll out services, customer service representatives need complete views across all lines of business. For example, insurance companies are adding services besides life, automobile, and homeowners insurance. Service representatives must access all relevant data. EII technology is ideal for this application.
- Federal Government. Certain laws and regulations make sharing information difficult. However, Homeland Security applications must integrate the data government agencies collect. Aggregating and copying data into a central repository won’t work. Agencies need to expose particular data elements, stipulate uses, and make specific data available to other agencies. EII meets this need.
EII-based data integration boasts a powerful new kind of enterprise technology that uses data federation. Visit: www.sybase.com/products/developmentintegration/avakieii or e-mail us at [email protected].
This article originally appeared in the issue of .