Building, Processing, and Deploying OLAP Cubes

After you’ve completed your OLAP cube by optionally adding translations, perspectives, actions, KPIs, calculations, and business intelligence logic, you’re ready to build and deploy it. During development, you typically use the default cube processing options so that you can quickly test and view the results. However, when you move to production, you probably want to use at least some of the myriad possible configuration settings. The reason for this is that during development you’ll often just use a small subset of data when processing your cube. In this situation, cube processing times will probably be short (usually measured in minutes or even seconds). Also, probably no one other than the developers will be accessing the cube, so if the cube is unavailable for browsing during frequent test full cube processing, then few people will be concerned.

The situation changes when you deploy your cube to production. There you could be working with massive amounts of data (some clients have cubes in the multiterabyte size) and hundreds or even millions of users who expect nearly constant access to the cubes. To prepare for these situations, you must tune the dimension and cube processing settings.