TM1 Forum

Hi All,

I have TI process which is running correctly .

My TI process : Test_TI and i renamed it to Test1_TI , In Test_TI iam creating the dimensions and cube and loading the file data into the cube.

After that i am exporting the same data into the CSV file and deleting that cube and dimensions from the system.

I am doing the samething in Test1_TI(I din't do any changes to the TI). But when iam running the Test1_TI it throws up and error saying cube need to be created.

I have Test_TI in the system only , when iam running my old TI it is executing sucessfully.

can anyone tell what is the reason.

Regards
Ashok

AshokReddy wrote: I am doing the samething in Test1_TI(I din't do any changes to the TI). But when iam running the Test1_TI it throws up and error saying cube need to be created.

So, Test1_TI is expecting a cube to exist that doesn't? It might help people if you provide the exact error message you're getting.

Are you using the output of GetProcessName() for anything? That's about the only thing I can see making a difference if the two TI processes really are the same.

To restate your problem
- You had a TI process that was working
- You saved that TI process with a different name, but made no other changes
- You ran the second process and it behaved differently to the first process

Is that correct? One of two things has happened here, the more likely is that you have accidentally or otherwise made some other sort of change. Such as commenting out the line that actually builds the cube. The other thing is that you have a bug. I think that's extremely unlikely.

I would suggest you delete Test1_TI, open up the original TI process and re-save it as Test1_TI again. If your problem is replicable then send it to IBM. However I've fairly sure you'll find that you have actually made some sort of other change.

As an alternative, you could copy and paste the prolog/metadata/data/epilog code from each TI process and put them side by side in Excel, then in the third column compare the two with the formula =trim(a1)=trim(b1). Anywhere that returns false means the code is different.

Martin

Martin Ryan wrote: As an alternative, you could copy and paste the prolog/metadata/data/epilog code from each TI process and put them side by side in Excel, then in the third column compare the two with the formula =trim(a1)=trim(b1). Anywhere that returns false means the code is different.

Just to expand on Martin's suggestion, there's also a very cool open source application called WinMerge (http://winmerge.org/) which will allow you to load up both .pro files and automatically find the differences.