Populate DSO

[Former Member]

Hi BI Experts!

We have written a source routine where we are populating an internal table with functional locations

(with some conditions). Then in field routines, we are checking that if fucnctional location of the souce field matches( from infocube) with what is present in internal table, then populate DSO accordingly.

When this is run with hardcoded functional locations through data trsnsfer process, the data in cube is matching exactly with the DSO. But when we run in general, the data is matching for some of the functional location and not for others.

In Data Transfer Process, we have 'Delta' as extraction mode.

Could you please suggest what could be the reason for this inconsistent behaviour.

Many Thansk,

Rohit

[Former Member · Accepted solution]

Hi

Just to be sure of the dataflow: CUBE -> DSO

in the DTP from Cube to DSO, you generate a table in the start routine with functional location.

When assigning value of Field 1, you make a look up in the internal table to check the functional location value, and depdending on the test result, you assign or not.

Are you sure your internal table is sorted ? Are you sure you make a binary search on it ? Are you sure of the key field in your DSO and on the asequence that the packages are treated to avoid overwritten data ?

PY

[Former Member · Accepted solution]

Yeah, we have binary search on the sorted internal table for functional location.

It seems that there is no overwrite of data as well as per the the primary keys of the DSO.

What we observed is that may be many records are not getting into Start routine from the Source package or few records are not coming at all in sourec package from the Cube .

when we run the data transfer process with fixed set of functional locations( one of the primary key of the cube), however the abap code is same but the DSO gets data which matches exactly with Cube.

This issue is really very tough and untraceable. Please guide.

Thanks,

Rohit

[Former Member · Accepted solution]

Hi

At this stage I can not be more helpful, except trying to debugg.

You could also have a look to the OSS note : 565086 which could help you findint out why you do not have all the records in that Datamart extraction...

Hope it will still help

PY

[Former Member · Accepted solution]

Thanks a lot.

We are still trying and let all know if any solution cracks.

Thanks,

Rohit