Developer's Toolkit

Blog archive

Architecting for Scalability

I had heard Pat Helland speak several times over the last couple of years, when he served Microsoft, and enjoyed his Megalopolis talk. That presentation compared an enterprise architecture to the evolving infrastructure of a city, putting the architect in the role of city planner and government official.

Well, Pat moved on to Amazon last year, but his talk this week on Architecting for Scalability was as entertaining and thought-provoking as any I had heard from his Microsoft days. One thought in particular that sticks in my mind was his assertion that it might be time to rethink some of the accepted wisdoms concerning data management.

For example, he notes that data normalization is not necessarily a good thing. Normalization, which manifests itself (in one way) as the requirement that data should only be changed in one location in the database, makes sense only if you're planning on changing that data, or, as Pat put it, executing UPDATE WHERE . . .

In many cases, that is a normal database activity. But he points out that is changing. First, offline storage (especially volatile storage) is cheap, enabling people to more easily save whole copies of databases for archive. Second, more businesses have to save database information as it is, rather than continually update it. Humorously, he notes that not only are business required not to update certain databases, but in the Sarbanes-Oxley era, doing so can be a felony (not so humorous, I suppose, if you are the one declared a felon).

But the point is that data management requirements are changing dramatically, and both architects and DBAs have to understand the implication to their jobs. In the case of architects, it has implications to how you go about architecting an application. Scalability has a different meaning when you have to treat data differently. In Pat's case, he notes that it becomes more important to not apply data across different business entities. It doesn't necessarily matter what database that data comes from; rather, it depends greatly on how it is used.

I don't normally get excited about data management, but I am in this case for two reasons. First, I work for a company (Progress) that produces a non-normalized database (although it can be validated for most of the behavior of the third normal form), so it was interesting to hear conventional wisdom being questioned in that regard. Second, it is always interesting to watch business requirements change and observe how both technology and skills adapt. We're still in the early stages here, so keep your eyes open to see how database technologies and best practices shift over the coming years to adapt to new business and legal standards.

Posted by Peter Varhol on 02/01/2006


comments powered by Disqus

Featured

  • Compare New GitHub Copilot Free Plan for Visual Studio/VS Code to Paid Plans

    The free plan restricts the number of completions, chat requests and access to AI models, being suitable for occasional users and small projects.

  • Diving Deep into .NET MAUI

    Ever since someone figured out that fiddling bits results in source code, developers have sought one codebase for all types of apps on all platforms, with Microsoft's latest attempt to further that effort being .NET MAUI.

  • Copilot AI Boosts Abound in New VS Code v1.96

    Microsoft improved on its new "Copilot Edit" functionality in the latest release of Visual Studio Code, v1.96, its open-source based code editor that has become the most popular in the world according to many surveys.

  • AdaBoost Regression Using C#

    Dr. James McCaffrey from Microsoft Research presents a complete end-to-end demonstration of the AdaBoost.R2 algorithm for regression problems (where the goal is to predict a single numeric value). The implementation follows the original source research paper closely, so you can use it as a guide for customization for specific scenarios.

  • Versioning and Documenting ASP.NET Core Services

    Building an API with ASP.NET Core is only half the job. If your API is going to live more than one release cycle, you're going to need to version it. If you have other people building clients for it, you're going to need to document it.

Subscribe on YouTube