Where I work, The data of sales is loaded into a "raw" table wich has a lot of entries (maybe 20.000.000). For internal reports, in the ERP, they use another tables that load from this "raw" tables and has the same data but sumarized.
They ask me to use this sumarized data for processing to avoid system overload, but i've found some errors on the sumarized data (ie. sometimes data is not updated by reason X).
What I would like to know, what should be a good practice for load data with QlikView,
a) Load data from the "raw" tables, using a lot of time/disc or
b) Load data from sumarized tables, having the probability of data mistakes?
