.NET Tips and Tricks

Blog archive

The Future of CA ERwin Data Modeling

My last blog post (CA ERwin and the Data Modeling World) was the start of an interview with Donna Burbank, the product marketing manager for Erwin, about the data modeling market. I finished the interview by asking about what CA Technologies sees in ERwin's future and the impact of ORM.

Peter Vogel: What's the strategic direction for CA ERwin Data Modeler?
Dana Burbank: One of our goals moving forward is continuing to evolve to support new data sources and platforms so that data management professionals can keep relying on CA ERwin as their "single source of truth" for data definitions and structural design. We're evaluating each of the use-cases carefully because the need is different for each role and technology. In some cases, integration with existing data sources makes the most sense, while in others it makes sense to model and manage the data natively in CA ERwin Data Modeler.

For example, a business intelligence developer might want to extract core data definitions from CA ERwin Data Modeler, and then design his or her application in their native BI tool -- that would be more of an import/export paradigm. In other cases, a developer may wish to model a new data source natively within CA ERwin -- a new cloud database, for example. We expect that CA ERwin will be used as the core hub for data-centric initiatives -- with CA ERwin being the common source for data definitions that can be reused in other projects: from BI, to MDM to application development, etc.

Visualization is the other key aspect to being this "single source of truth" for data assets. One of the reasons that data modeling has become so common is that it helps people visualize complex data structures. Visualization options have changed dramatically from the traditional data models in the 1970s and 80s. You'll see some next-generation ways to graphically show data models in upcoming releases.

We're also seeing personnel with a broader range of responsibilities getting involved in the data modeling process -- from business users and executive sponsors to technical staff from a variety of backgrounds. To address these audiences, we're looking at non-traditional ways of showing data models: from the simple but familiar spreadsheet view of business users, to BI-style reporting, to more alternative ways of sharing the core metadata and definitions that are managed in ERwin. This is key, because when more people across an organization use the models, more value is drawn from reuse and standardization, and ultimately costs are lowered and production time is decreased.

PV: What, if any, impact do you see ORM tools having on the market that CA ERwin Data Modeler is part of?
DB: Although ORM is a topic of discussion in academic circles and at conferences, we're not seeing a huge demand for ORM from customers planning to use it for their core design projects.

The strength we see for ORM is for high-level or business-level data modeling, which is known as conceptual data modeling to many data architects. ORM has a nice way of displaying core business facts, and its emphasis on showing sample data sets within the model can be an intuitive way for business users to understand the data relationships being discussed.

CA ERwin Data Modeler continues to focus on the Information Engineering and IDEF notations, which can also show conceptual data models, but can easily translate to the logical, physical and implementation levels, so that these conceptual artifacts can be leveraged throughout the design process.

Posted by Peter Vogel on 12/23/2010

comments powered by Disqus


Subscribe on YouTube