TM1 Forum

Many many thanks, again, to Alan for booting me in the rear and making me learn to love Turbo Integrator. I am getting pretty good at it.

I have been working with TM1 Vers 6 and am converting our system to Vers 9.4. All the data has been transferred over and I am currently writing processes to duplicate the old Vers 6 functionality.

We have accounting data in another database. In the old system, the Data Control job would query the accounting data and create a cma file. Then, it would read the cma file into the cube using a map sheet in Excel.

On the new Vers 9 side, I have created a query of the accounting database that sends the data into Excel, in a table. I then have various formulas that manipulate the data (a little is necessary), then uses DBSS functions to write it in the appropriate cube. This works.

My question is: Am I doing it efficiently? Should I, instead, query the accounting database, create the cma file then use TI to read the cma file into the cube? There is some minor manipulation necessary, easily done with Excel formulas.

Why go through the step of creating a temporary file? Just use the query as your data source. If you need to modify the results you can create a new variable and assign a formula to it that manipulates the number into what you want it to be. It's pretty easy to do.

Depending on what your database is Tomok is right. TI can connect directly to the datasource via ODBC. The TI should then be able to do the manipulation you are currently doing with the excel file, cutting out all manual intervention (and hence margin for error) and should speed up the start to finish process.

I found this message while searching for something else.

Why go through the step of creating a temporary file? Just use the query as your data source

That works perfectly, and thanks.