Question about using Chore to update dimension/cube

[lynnsong986]

Hello,

It seems that I need to rename the updated CSV files and take the browser to the renamed file for the TI process to recognize changes made in that file to update a dimension or a cube. This doesn’t make sense to me - if I create new chore that updates the dimension or the cube every 7 days automatically, how can TI pick up the changes in the CSV files because the path/file name would be the same?

Thanks!
Absolutely a beginner

[Wim Gielis]

Hi,

You might be a beginner, in fact I guess that experienced people will also struggle with this one.

As it appears from my testing, you need to store/upload the file in the 'model_upload' folder, which is part of the TM1 data directory.
Actually, when you browse to a file, PA will upload a copy of the file and store it in that folder.
That file has no direct connection with your local/first file, so updating the first file will not let you see the details in the cube / dimension / subset / ...

02.png (40.26 KiB) Viewed 7478 times

I did not see this behavior explained in the documentation - but that's hardly a surprise:
https://www.ibm.com/support/knowledgece ... ource.html

I noticed this by opening the *.pro file for the process from the TM1 data directory (or opening up the process with Architect).

Unless someone tells us it's possible, this will mean that we cannot have our traditional input folders (Data, Metadata, ..., folders by dimension, etc.)

Best regards,

Wim

[lynnsong986]

It’s great to know! I can stop searching for answers now...

Thank you so much for answering!!

[Wim Gielis]

I did not see this behavior explained in the documentation - but that's hardly a surprise:
https://www.ibm.com/support/knowledgece ... ource.html

Unless someone tells us it's possible, this will mean that we cannot have our traditional input folders (Data, Metadata, ..., folders by dimension, etc.)

Let me answer my own question, I found it in IBM documentation by the way.

This works in Architect/Perspectives (PM not tested but probably as well):

PA paths.png (77.64 KiB) Viewed 7459 times

The screenshot is taken from a Partner platform on the cloud provided by IBM.

Wim

[lynnsong986]

Thank you so very much for your time on this! You are totally awesome!

[tomok]

The model_upload directory only exists of you have used PAW to create a TI process. We are on the IBM cloud and we do not have any such directory in our data directory. I'm assuming it is because we have not used PAW for anything yet. We had to create our own directory for storing text files to be used as a data source and we have to use either FTP or Cognos Command Center to upload anything to our TM1 server.

[Wim Gielis]

The model_upload directory only exists of you have used PAW to create a TI process.

Correct. I wonder when this limitation of PAW will be removed.

Earlier post: https://www.tm1forum.com/viewtopic.php? ... 192#p70085

For a video on which paths to use (they only discuss the relative path option), refer to https://www.youtube.com/watch?v=qoFW--H ... e=youtu.be

[Wim Gielis]

we have to use either FTP or Cognos Command Center to upload anything to our TM1 server.

I see that the remote desktop connection allows making your local resources (C, D, drives, ...) available in the remote server. That's probably easier than FTP and similar tools.