All Forums General
Milan 2 posts Joined 11/13
13 Dec 2013
GROUP BY CUBE or ROLLUP how to use it?

This is more general query from beginner and not pariticular problem or issue. 
I am experimenting with GROUP BY CUBE and ROLLUP functions on Teradata,I understnad basic concept, but I am missing something important.
We have CUBE created, but how to utilize it? Creating CUBE using GROUP BY CUBE is not an OLAP cube, Am I right? It is just somthing like agregated output stored as a normal table.
Is there any way or tool which actually can utilize such CUBE?
Are there any tools which can connect and utilized what GROUB BY CUBE has created?
I am still thinking it is just a table and thus seems to me that GROUP BY CYBE function is not particularly usefull. OLAP CUBE should allow you to rapidly create diferent output views.
GROUP BY CUBE has everything nicely agregated, but so far I can just use regular SQL to query it and it is far from rapid view modeling. On contrary you must know structure of such table very well get out what you need.
I know there is something like TD OLAP Conneector, but it probably not utilizing GROUP BY CUBE tables?

Raja_KT 1246 posts Joined 07/09
13 Dec 2013

Hi Milan,
I have never experimented GROUP BY CUBE. Teradata has Aggregate designer and OLAP connector etc. 
You can have a look at the tool and see and please share your experiment and thought.

Raja K Thaw
My wiki:
Street Children suffer not by their fault. We can help them if we want.

M.Saeed Khurram 544 posts Joined 09/12
14 Dec 2013

Hi Millan,
The first thing is what is a cube? A cube is simply used for multidimentional processing. it is also called OLAP Cube, and Teradata has extended its Group by cluase to easily handle all these ROLLUP and CUBE creation withing SQL. I have not read anywhere that a Group By Cube is stored as table, Can you please provide some reference how did you deduced that?
If you want to use this Group By Cube in any thirdparty reporting tool, I think you can use it withing your views. Also the main purpose of the Group By Clause is to perform analysis in a regular SQL Query, to create and store a Cube just SSAS does, Teradata might provide some other process.
Have you tried it using in you sementic layer and then access using some third party tool?


Milan 2 posts Joined 11/13
14 Dec 2013

Thank for answers guys,
To Raja, yes I have read few articles regarding TD OLAP connector and also about teraedata agregate designer. My problem is that I am working in a corporation, chance to get some lycences is very low. However those AJIs looks very cool, I will try to use them.
to Khurram, that exactly my question if there is a tool which can utilize GROUP BY CUBE creation.
Regarding CUBE stored as regular table, I did it that way and want to know if there  is other way. 
I need some examples how it can work something like: we have used GORUP BY CUBE, then we store it like this ant that, then we connect to it with some OLAP conenctor, ... 

You must sign in to leave a comment.