Hi All,
I am new to Powerplay transformer. Since i have got some experience on the Other OLAP tool SSAS (SQL SERVER ANALYSIS SERVICES), I didnt find this tough enough to learn. I have learnt some primitive steps on how to build the cube in Powerplay but that is not enough for designing big cubes and not well versed on the design considerations. Sure i hope i will gain fluentzy as the day proceeds.
If you are worked in Powerplay Transformer, you may wonder why we need 2 formats of model files (.mdl and .py?) like me. I found out what are the advantages and dis-advantages of 2 model file formats.
If you come across the Cognos Transformer a OLAP tool to model your cubes, you will be familiar with two file types .pyi and .mdl. The base model for the Cube will be saved in these formats. I wonder why we need two formats to save the model? And what is the major difference between these 2 formats.
Today as I browsed through the help files of the cognos documentation, I come across a File named pp_trns_mdl.pdf (Transformer MDL Reference) which explains all about the MDL. I assure you that I not even looked into those files after I install cognos 2 weeks back .lol!
I just skimmed thru the first few pages of that document; I was absorbed by the answer which I was looking for? The first chapter itself explains about the difference between two formats.
Now let’s look at the different features of those two file type (which is in the File I mentioned):
.mdl – Features
1) You can open the Models saved in the .mdl format in the User interface or in the text editor.
2) The .mdl format is the plain text representation of the model that is compatible between the versions of Transformer.
3) It loads more slowly than .py? Format because it creates all of the transformer objects when it loads.
4) Models saved in the .mdl format are machine independent.
.py? – Features
1) The py? Format is the binary representation of the model.
2) It is not compatible between the versions of the Transformer.
3) It loads faster than .mdl format because the Transformer Objects are not recreated when the model reloads.
4) In general .py? Models are larger than .mdl files.
5) As you edit the model, the size of the associated .py? file increases because it maintains the information on the operations performed during model editing.
6) Transformer uses this information in the variety of contexts such as Client-Server Operations and incremental updates.
Today as I browsed through the help files of the cognos documentation, I come across a File named pp_trns_mdl.pdf (Transformer MDL Reference) which explains all about the MDL. I assure you that I not even looked into those files after I install cognos 2 weeks back .lol!
I just skimmed thru the first few pages of that document; I was absorbed by the answer which I was looking for? The first chapter itself explains about the difference between two formats.
Now let’s look at the different features of those two file type (which is in the File I mentioned):
.mdl – Features
1) You can open the Models saved in the .mdl format in the User interface or in the text editor.
2) The .mdl format is the plain text representation of the model that is compatible between the versions of Transformer.
3) It loads more slowly than .py? Format because it creates all of the transformer objects when it loads.
4) Models saved in the .mdl format are machine independent.
.py? – Features
1) The py? Format is the binary representation of the model.
2) It is not compatible between the versions of the Transformer.
3) It loads faster than .mdl format because the Transformer Objects are not recreated when the model reloads.
4) In general .py? Models are larger than .mdl files.
5) As you edit the model, the size of the associated .py? file increases because it maintains the information on the operations performed during model editing.
6) Transformer uses this information in the variety of contexts such as Client-Server Operations and incremental updates.
Tip provided in the same help file: Storing model operations can cause model fragmentation within the binary model. To eliminate this fragmentation save .py? models periodically in .mdl format.
Best practice: It is a best practice to save your model files both in .mdl and .py? formats.
bye..
No comments:
Post a Comment