Hi,

Customer has loaded dimension once in their production environment.
We know their are new dimension values that have been created by source systems.

What would be the best practice to load these new dimension values?

Load only the new records and make sure the "Delete if not new" check box is unchecked?

Is it better to reload dimensions entirely? Full refreshes?

I am trying to get a decision with the customer on best approaches when loading dimension data on a weekly basis.

Merci,

A plus tard.

Unlike the Measure Loader, there is no such thing as a "Full Refresh" on a Dimension Load.

If you wanted to simulate a "Full Refresh" on a Dimension, you would have to delete the Dimension and then re-create and re-load it. Needless to say, this would not be a good thing to do, since the surrogate keys (which are the hidden database keys that link the Measures to the Dimension values) would all change, and any Measure records using that Dimension would be invalid in the PMF database. In fact, it is for this very reason that PMF does not allow you to delete a Dimension if there are any Measures using that Dimension.

So the only option available to you on a Dimension Load is "Delete if not in new". Before going into details, it should be noted that there is no harm done by leaving this box un-checked (in other words leave all previously loaded Dimension values in PMF even if they are no longer in the source data). The worst thing that can happen is that you will see obsolete Dimension values on the Dimension Trees and drop down list boxes

So now, the decision to do a "Delete if not in new" depends on the nature of your business requirements. What does it mean that a Dimension value in PMF from a previous Dimension Load is no longer in your source data? Was there some sort of re-organization or change in your source data? Are there any source Measure records that were using this Dimension value and if so what happened to them? Are there any records now in PMF that are using this Dimension value and, if so, what should happen to them?

Once you've gone through this analysis, the answer should be clear; but if not, don't hesitate to ask follow up questions!

EricH