CODEWITHSUNDEEP
iccube
John
John

Reputation: 313

icCube full reload of dimension tables

I think I need some clarification on how the full reload works as an incremental load strategy for dimension tables.

I have a Dimension with multiple attributes (represented as a multiLevelDimension with multiple hierarchies). The dimension data is loaded as a CSV and has the incremental load strategy set to full reload. I've loaded a number of partitions into my cube and then changed one of the attributes for the dimension. When I now query the cube, I'm getting results in the image below.

Here the InjectorName is the default hierarchy and FuelType is an attribute of the Injector (changed from Coal to Gas). Both InjectorName and FuelType are loaded from a single CSV file. The Case dimension in the columns of the pivot table corresponds to the partitions.

I thought that the full reload strategy would override the old association of Injector with FuelType = Coal and I would only have a single occurrence with FuelType = Gas but as shown below, both associations appear.

ChangingAttributes

Upvotes: 1

Views: 42

Answers (1)

Marc Polizzi
Marc Polizzi

Reputation: 9375

While performing an incremental load icCube does not update/delete existing members; only new members can be added. That is why you are still seeing the "Coal" member in the pivot table.

The strategy "Full Reload" applies to the loading of the data from the table and says in your use case that the table is fully reloaded to detect the new members.

Hope that helps.

Upvotes: 1

Related Questions