Adding IQD Datasources???

[mjcotter]

Check this out. I have a transactional datasource that
I had to break out by month. I have 36 months in my database, so now I have 36 transactional datasources.

I am trying to insert these 36 datasources into my Transformer model. It is taking a ridiculous amount of time to insert 1 datasource. I clocked it at about 2 hours just to insert each datasource.

Can anybody help??? Why does Cognos take so long to insert each datasource. I would expect it to take a couple of minutes tops. Not a couple of hours.

I will soon have a Transformer model which will have 108 IQD datasources. That means it will take 27 days to insert each datasource!!! That is crazy if you ask me. Did Cognos really design such an inefficient tool? I've tried their support line, but they've been no help.

Thanks!

 

[Guest_imported]

When you say add to the model do you mean building the cube? If so you should check and see how much time is used to read the datasource using some other query tool.

 

[mjcotter]

I mean bringing the datasource into the model for the purpose of using it to build a cube. I'm not even at the point where I am trying to build the cube.

Cognos Support tell me it is because Transformer goes through all of the categories assigning them to the new datasource. They think I should do a Clean House, but that is not really the best solution because we have some categories that we created in the model that will get blown away on a Clean House.

It's too bad there isn't an option in an .INI file where you can tell Transformer not to do the category assigning.

 

[WondersWhy]

Do your separate datasources have similar or identical column names? Tony T. in Mpls.

 

[mjcotter]

They all have the same exact columns. The only difference between them is a filter, which basically filters each data source by month. We needed to do it this way because the resultsets were too large.

 

[WondersWhy]

I guess I cannot really help here if you are not able to get over the large resultsets problem. I can only imagine that if you can pull the data in as one table (Access Query, IQD file, whatever) things would be fine. The maximum records I pull into Cognos is about 130000 records. Maybe that is not so big. I do not know of any input limitations because I haven't hit any yet. I suppose if your ever using Excel as a datasource you might have what, 65,000 as your limit?? What part of your process presents the size limit? It seems that if you could solve that problem you might be OK. Is there any way to purge irrelevant or ancient data somewhere upstream of your constraint?