If you are unable to create a new account, please email support@bspsoftware.com

 

News:

MetaManager - Administrative Tools for IBM Cognos
Pricing starting at $2,100
Download Now    Learn More

Main Menu

D-List Update using Import Link

Started by andy_acacia, 04 Oct 2010 12:20:06 PM

Previous topic - Next topic

andy_acacia

This one has me stumped!

I am using an import link to update a D-List. The D-List happens to be a cost centre list in a profit centre hierarchy.

The import runs off a D-Cube containing all the hierarchy data - ie. 1st Dimension - Cost Centres, 2nd Dimension containing D-List formatted Profit Centres, Functions, Company Code.....
The D-Cube is fully populated. No cells have missing data.

When I run the D-List update it brings in ALL the Cost Centres, but not ALL the profit centre hierarchy items.
So I have a list of cost centres rolling up to profit centres and then a bunch of cost centres that are not allocated to any profit centre! This despite the cost centre clearly matching a profit centre in the D-Cube that is used for the update.

I have used update links millions of times without an issue. What could be going wrong??? Is this a Marxist conspiracy or am I doing something really dumb?

andy_acacia

Had an idea - I reduced the number of cost centres to test my theory.

By reducing the number of cost centres, suddenly the update works! Hierarchy Profit Centres that previously did not import are now visible??
I see no reference in Cognos documentation referring to scale issues on update links.

I still need to run the update on the full Cost Centre list so any help still appreciated.


skinners666

Sounds very interesting. How large is the import? Is it possible to recreate this with a library you can send so I can have a look?

andy_acacia

Thanks skinner.

you got an email address that I can send the LIB to?
Alternative I could post the folder here but not sure if there is an easy way to do this?

skinners666


skinners666

OK - Ive had a look ...

You have set unique names on the dlist to be the first 10 characters.
Some of the parent items are not unique in thefirst 10 characters in the cube.
For example 271102 IT infra and 271102 IT base.
The first item in the dlist not placed in a hierarchy is called 2000251000.
According to the cube its parent is 271102 IT base.
But there already exists in the dlist a parent called 271102 IT infra (item #576) and so the parent cannot be imported to the dlist as it has a non-unique name.

Hopefully that makes sense ... basically its caused by the uniqueness id on the first 10 chars ...

andy_acacia

Skinners
my apologies, a thanks you is very very much overdue!
I have only just spotted that I never responded.

Thanks again