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

mdl / pyj difference

Started by nfldbunker, 27 Jan 2010 07:33:54 AM

Previous topic - Next topic

nfldbunker

hi

I am really new to cubes and transformer and I need to modify a transformer cube. Currently we have the pyj file and the mdl files(backup).

if i want to modify the transformer model which file do i open and modiufy(pyj or mdl). Do i modify the pyj and then save it and after complete resave to a mdl version for backup.

This is in our production environment so i am a little hestiatant to comeplete.

thanks

bunker

 

redmist

There are few differences between the two formats and they are explained in the documentation.

Couple of key points for me were
I generally build the model in .pyj format. This is a binary format and so if you wanted to move between different platforms such as windows/Unix this would be one to use. When you save the model in .pyj it is saved after compiling and so its size would be pretty big and it would take some time to open.
.mdl files are not compiled so their size will be much smaller that the pyj version and it would open faster too. But the advantage here with pyj is that cube build would be faster as the mdl file has to first compile the data and then start the cube build.

Best practice is to use the pyj as your main model and back it up to mdl frequently. This is due to the fact that sometimes the categories in pyj gets corrupted. Backing it up to mdl will ensure that you have a clean backup.

In your case open your pyj model and do save as mdl. Now save the mdl back to pyj format.

You might want to keep a backup of your existing model before doing this