binary incremental load

[#1]

Hi does anyone have knowledge how to setup binary incremental load? I have read that it is much faster then regular qvd incremental load.

So instead of keeping historic data in QVD it only contains the incremental loaded data.

Hope someone can help

Thanks in advance

[swuehl]

Not sure what you mean with binary incremental load, you can only perform a single binary load in your script.

Could you reference where you have read about this, so to give context?

[#3]

Yes of cause :

http://www.qlikviewaddict.com/2012/02/power-of-binary.html

I have a problem with that when i reduce and distribute my qlikview document it takes over 8 hours (and the file only consists of 50 mil rows) So it really shouldnt take that long.

But i cannot seem to figure out why its so slow

[Colin-Albert]

The binary command loads all data from one QVW into another.

To add the delta data for an incremental load, you would need to be able to identify the most recent records in the tables loaded in the binary command and then append new data to those tables in your load script.

This should be straightforward if the delta simply consists of new data, but will become more complex if you need to handle deleted / modified records and slowly changing dimensions.

[Colin-Albert]

Which version of QV server are you using? Ver 11.0 changes in the way documents are opened for distribution which may affect your performance.

[#6]

I use Version 11.20

[Peter_Cammaert]

In the affected document, can you pinpoint where exactly this excessive amount of time is wasted? Or is it the Reduce step itself? (in that case the document log should show speedy execution, but the task log will show a big delay somewhere)