is it possible to use Qlikview only as a query tool without importing any data, in order to maintain the logics and the relationships of the data source. Fe Qlikview connects to a SSAS cube and read the related measures: why rebuilding the logic of a measure in QV? Microsoft Excel connects to a SQL Server source or a SSAS source without importing any data and queries the source by a pivot table.
Well you do have the option of having a hybrid approach with QlikView 11.2 Direct Discovery.
And then loading the dimensions into memory is nothing different than other tools would do. They just call it caching or something else. There are no tools that does not import ANY data - Data needs to be in memory to be processed in all standard computer systems. The logic does not need to be rebuilt if you already have a good logic in your source system there is less work to be done in QlikView. QlikView is very flexible in the way that it does not force you to build cubes. Most of the times however a star-modeling approach is the best approach in QlikView too. There is nothing in QlikView that prevents you from applying best-practices that are well-documented by for instance Kimball.
QlikView comes from a very different tradition which actually was quite revolutionary and definitely not considered best-practice. Nowadays however it seems like the inventions and innovations in QlikView are accepted as best practice. Look for instance at what Microsoft calls Tabular Modeling which they introduced I think in SQL Server Analysis Services 2012 - it has many similarities with how QlikView has done it since 1994.
My goal is to understand if it is possible to use Qlikview to show data contained into a SSAS cube, in particular into a Tabular model. This model has some dimensions and some fact tables, but also a set of measures. Some of these measures are customized, not are simply a SUM or an AVG of an amount. To use Excel to show the tabular data I need to create a connection against the Tabular model and create a pivot table, without importing any data; when I filter for certain dimension the implemented logics for the used measures allow to show the correct value for them. Now I'd like to use Qlikview and not Excel as a presentation tool, but I don't want to rebuild in QV the implemented calculations for the Tabular measures; I need to avoid to do a further effort to develop again these measures in QV.
So if importing SSAS measures in QV means importing data and the related logics is a good thing otherwise no. I need to preserve the dynamic behaviour and the correctness of the source measures.
I hope to explain better the issue. So, which is the best approach to follow in QV in order to achieve my goal?
You will connect to a SSAS Tabular model the same way as you connect to an SSAS Dimensional model (OLAP). So you could either enable distributed queries and then write MDX-queries yourself that can be written in the SQL statement in a QlikView load script or you would have to rely on a third-party solution. You can also find the OLAPExtensions on CodePlex that will enable you to send T-SQL queries directly to any OLAP server both dimensional and tabular.
Hi Petter, but you don't say to me in an explicit manner which is the best approach to follow in QV in order to connect a SSAS Tabular model avoiding to lose the dynamic behaviour of a Tabular measure and so avoiding to develop again it.
Do I need to the Direct Discovery feature? If yes, I've tried to write the script to read a Tabular model by the Direct Discovery feature but unsuccessfully. I've found that in order to read a SSAS Tabular model I need to use a DAX query and not a MDX query as for a SSAS multidimensional model. I've used with success a DAX query to load a measure in QV in the traditional manner.
Do you have any examples about writing the script that uses the DIRECT syntax to read a SSAS Tabular model, please? Many thanks
Maybe you should have a look at the DataForce Solutions product named O-Delta an "OLAP to QlikView" solution. It might be something in that direction that you are looking for...
I don't know more about it than I have read on their web site - you should contact them for more information as you will find their contact details there:
