The Microsoft® Data Warehouse Toolkit: With SQL Server™ 2005 and the Microsoft® Business Intelligence Toolset

17.1. Making the Case For (and Against) Real-Time Data

Real-time data sounds so cool. Who would not want to have data that's as fresh and current as possible? And surely we, as professionals, want all enterprise data to be conformed and consistent? What better way than to populate the DW/BI system in real time?

As we discuss in this section, it may sound cool, but it's not so easy.

17.1.1. What Makes Delivering Real-Time Data Hard?

The hardest things about delivering real-time data are, as usual, related to people rather than to technology. Nonetheless, there are some technical challenges, too. Microsoft offers many interesting features to help you with the technical challenges.

First, the people-based challenges. One of the biggest issues for the DW/BI team is to meet both the demands of those who want real-time data, and those who most decidedly do not. You may think that pushing the latency closer to real time would be a win for everyone. Surely, in the absence of any cost of delivering data, we'd all prefer to have real-time data?

Actually, no. Anyone who's trying to develop a non-trivial analysis knows that you need to work on a static dataset, where the numbers aren't changing from moment to moment and query to query. If you try to make these analysts work against a dynamic database, they'll copy a bunch of data to a personal computer: exactly the kind of behavior you're probably hoping to stem with your DW/BI project.

A second person-related problem has to do with the ...