Continue in 2 seconds

Data Modeling Over a Barrel – Using System Architect to Model Data Structures at Interbrew UK Ltd.

  • September 01 2004, 1:00am EDT

CATEGORY: Data administration, management, modeling and database performance.

REVIEWER: Paul Drake, data architecture manager for Interbrew UK Ltd.

BACKGROUND: Interbrew UK Ltd. (IUK) is the UK arm of Interbrew, a global brewing company whose origins date back to 1366. Interbrew has a portfolio of more than 200 brands, including Beck's, Stella Artois, Leffe, Hoegaarden, Labatt and Rolling Rock.

PLATFORMS: Intel-Based PCs running Windows XP Professional and Windows 2000 Server.

PROBLEM SOLVED: IUK has used the Popkin System Architect (SA) toolset to produce logical data models for all of its new developments for more than five years. More recently, a business information team has been created to rationalize IUK's reporting and business intelligence. They are creating an operational data store (ODS) to hold data consolidated from multiple sources. To ensure that the ODS provides a well-documented business view of IUK data, a corporate data model (CDM) was constructed using SA.

PRODUCT FUNCTIONALITY: IUK currently only uses SA's data modeling functionality. However, this functionality alone justifies the cost. SA provides an environment where you can draw entity-relationship models of business data, to which you can add rich textual detail. The ODS is being built on an on-demand basis. Using SA, we initially built a high-level CDM showing basic entities, attributes and relationships. The tool removed lots of the pain of the creation and maintenance of the model by propagating foreign keys, spell checking descriptions, etc. Once we had a working CDM with enough detail to proceed, we used SA to create a physical data model and then to generate the database. This gave us our first-cut ODS, with its logical counterpart fully documented with business descriptions. As a request for a new report comes in, the requirements are investigated. A new diagram is created which is a subset of the CDM. Existing required entities are included from the CDM. Any new entities and attributes that are created here are automatically added to the CDM. This feature is very powerful. It allows small, easy-to-understand models to be produced without worrying about keeping them in sync with the CDM (as they are already implicitly part of the CDM). Any changes made to the CDM can then be fed down to the ODS database. Because this keeps the CDM and ODS in sync, the documentation is always up to date.

STRENGTHS: SA's major strength is its flexibility. Because it is extensible, if there is data that SA doesn't support, you can easily add it. Also, it includes VBA, so you can write your own code to enforce local standards, produce formatted reports, etc. The ability to easily produce subsets of a data model is not unique to SA, but it is well-implemented.

WEAKNESSES: The report writer provided with SA is very basic, although the recently added ability to produce reports in XML format has improved their usability. However, the VBA object model is powerful and can be used to create complex reports, albeit with some effort. The GUI is often criticized by the user base as being old-fashioned, although I personally find it quite easy to use.

SELECTION CRITERIA: Select SE and ERwin were considered, but SA's flexibility and ability to support multiple methodologies (UML, process modeling, etc.) made the decision easy. (We intend to extend the use of SA in the future.)

DELIVERABLES: SA produces rich logical and physical data models and DDL to create SQL Server databases.

VENDOR SUPPORT: Popkin is one of those rare software companies that is more interested in helping you succeed long-term than they are in making a quick buck. Their staff - from the help desk to the consultants - is knowledgeable and friendly. The User Group ( is independent of, and highly regarded by, Popkin. I can honestly say that Popkin is one of my favorite suppliers.

DOCUMENTATION: The documentation is adequate but not fantastic. Finding answers can be a challenge unless you know the terminology. However, the tool is easy to use once the basics are grasped. The biggest problem lies in understanding the methodologies, and it is not Popkin's responsibility to teach that. The extensibility documents have been poor, but are improving with each iteration.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access