Dynamic Cubes

[esar]

Hi,
I am trying to have a dimension added to a report built out of a Dynamic Cube. The dimension is a flag indicator coming from Fact table. So joins were not required in the Dynamic Cube modelling. But when I use this flag in report Slicer, it throws an error XQE-PLN-0091 Cross joins (between query subjects: ...........) are not permitted..

Have anyone faced this before. Any help would be appreciated.

note: I even tried the governer settings in FM to Allow Cross Product.

[MFGF]

Hi,

Dynamic cubes have dimensions as dimensions and facts as measures. You cannot have a dimension derived from a fact. You could perhaps create a database view that makes the new data appear as a dimension linking to the fact and model it as such in cube designer?

Cheers!

MF.

[esar]

Hi MF,

Thanks for your quick reply.

I agree to your design to create a view on top of it. But I was refering to the following link in IBM which says, Dynamic cubes supports degenerate Dimensions. It also clearly depicts my scenario. Any opinion on this ?

http://pic.dhe.ibm.com/infocenter/cbi/v10r2m0/index.jsp?topic=%2Fcom.ibm.swg.ba.cognos.ug_cog_rlp.10.2.0.doc%2Fc_cmddimension.html

[MFGF]

Well, that's news to me! :) I haven't seen anything in Cube Designer to allow this, but that doesn't necessarily mean it's not there - just that this muppet hasn't come across it :)

MF.

[Francis aka khayman]

trying to find links how it can be accomplished... not having any success so far :(

[esar]

Am trying Hard too.. There must be some way around; else why would Cognos provide Degenerated Dimensions as an working option. I will try to open a ticket with IBM. Hope they dont reply back asking me to upgrade to 10.2.1 from 10.2.0.

Any Other Expert opinion? My last option would be to go with MFGF's approach.
:-[

[MFGF]

I did a bit of digging. Looks like degenerate dimensions are supported to assist in the definition of aggregate cubes but this doesn't necessarily mean they can be used for a normal cube.

Take a look at the Dynamic Cubes Redbook - Chapter 4, section 4.9 to see what I mean.

http://www.redbooks.ibm.com/redbooks/pdfs/sg248064.pdf

I still think my original ramblings were the way to go :)

MF.

[esar]

Got the below response from IBM Support after sending them the component list.

"Looking at the cmplst files it looks like you are running on the GA release of 10.2 and it may be that you are running across a defect with how degenerated dimensions behave in that version. Once such defect is:
http://www-01.ibm.com/support/docview.wss?uid=swg1PM83553"

Responded to request and sent them my model file for further investigation.

[MFGF]

Ha! Looks like the one thing you hoped they wouldn't say is what they are saying :)

Oh well. I'm not convinced that upgrading will allow you to do what you need to, but we can hope :)

MF.

[esar]

Weirdly.. This model works in my QA instance.. This issue persists only in Production..