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

Finding Reports based on a specific Join in FM

Started by Ravisha, 22 Dec 2014 07:57:39 PM

Previous topic - Next topic

Ravisha

Hello Cognoise,

I came across an issue for which I need your help. I need to find all the reports that uses a specific join in FM model.

For example:

Assume that Table A has a relationship with Table B. I modified the model by deleting the join between Table A and Table B. Now, I need to find all the reports that might have an impact.

FYI, I tried using 'Analyze Publish Impact', and it says "there are no changes made that might have an impact"  :o

I would appreciate it if anyone of you could guide me through the process or solution.

Thanks in advance

CognosPaul

I believe that search looks for objects from the tables. I recommend opening an enhancement request with IBM to better support joins.

Why not make a change to Table A, and get a list of reports impacted, and then do the same for Table B, and find all entries in common. This should find reports that use both tables, but I'm not sure what would happen if your model is a snowflake (Table C connected to A and D connected to B).

Ravisha

Hello Paul,

Thank you so much for your quick response. I really appreciate your time and support.

Yes, you are absolutely correct. Table C -> Table A -> Table B <- Table D.  The model is highly Snowflake and is completely relied on OLTP DB.

I only way that I could figure  out is after making changes to the Dev model and pointing the Datasource connection to Prod DB, copy all the published reports and point it to Package_Dev and perform a Regression test on it and then I'll compare it to the outputs of already published reports pointing to Package_Prod. I know that it's a tedious process but that's the only idea I could figure out. I would really appreciate any guidance or help on this.

Thanks a lot sir !