Missing records in Data Target

[Former Member]

Hi,

We have a custom cube with fairly straight forward update rule and transfer rule. The request id status is all green and the number of records in Selected, Received, PSA, Converted, Updated, and New is 8863 records.

But in list cube for that particular request id, we get only 7993 records. No further filter in the selection. Has anybody faced this kind of situation or have any clue as what has happened? Any help is highly appreciated.

Thanks in Advance,

Praveen

[Former Member]

You could potentially bump into such a situation if you are using datasources / infosources like 2LIS_12_VCITM, 2LIS_11_VAITM, ...... which were installed before 2.0B sp 7 and never implemented note 333492 later.

Please check this note & see whether it applies to you.

Kumar Gudiseva.

[Former Member]

Hi,

We have figured out the issue. The records are getting loaded into the Fact table but with different DIMID for the Data Package Dimension. For the same Request ID we found two DIMIDs as there seems to be a different recordtype (0RECORDTP). We are now collapsing the requests so that the records are available for the Reporting.

Thanks everyone for your postings.

Praveen

[former_member184494]

praveen,

There is a possibility that some of the records could have got summed up , hence the reason for lower number of records. For finding out the same:

1. Copy contents from PSA and from cube and compare - most likely in Excel.

2. Create an ODS - check contents with the same rules...

3. Also check in RSA3 for number of records...

Arun

[Former Member]

Hi Arun Varadarajan

Thanks for your Reply.

The records are unique and could not have got summed up.

We created another temporary cube to load them and all got loaded as 8863 records. The data load was Delta upload and loaded a month back and we noticed this issue now.

Thanks,

Praveen

[Former Member]

Hi,

Any more clues or suggestions.

Thanks,

Praveen

[former_member200327]

Hi Praveen,

I guess you've loaded this new cube from same PSA, not from R/3. Now you have two cubes with supposedly same data. You can create a multicube on top of them and run a report to find records that are missing. Those records should provide some clues to what happened.

If you don't want (or can't) create a multicube you can create same Queries for both cubes and compare the results.

Please let us know what you found.

Thanks,

Gersh

[Former Member]

Hi Gersh,

Thanks for your reply posting.

Yes. To test that data request alone, we have created a temporary cube and loaded from PSA.

We can not create another cube only for those records as the orignial cubes are being used for almost last 2 years.

We are still trying to find out as to what happens to the missing records to solve the issue permanantly

[former_member200327]

Hi Praveen,

If you "old" cube has other requests you can put a request number in selection criteria in the query for that cube in the multicube.

Message was edited by: Gersh Voldman

[former_member200327]

Hi Praveen,

You can check how many records you really have in the cube by going to DB table /BIC/F<cube name> (if you didn't compress the cube). Number of records in this table is real number of records in the cube.

I can think of at least two reasons why those two numbers are different:

1. You have multiple records with same combination of characteristics in one load. They counted as separate records by the load program, but they are combined in one record when written to the cube.

2. You selected 'DB aggregation' check mark in Listcube. This will show records with same combination of characteristics (especially if you selected not all characteristics) as one record with summarized key figures.

Hope this helps.

Regards,

Gersh

[Former Member]

Hi Gersh Voldman,

Thanks for your Reply.

Number of records in DB table /BIC/F<cube name> also returns only 7993 records instead of 8863 records.

The records are unique and not combination and hence should have loaded the same number of records but not so.

We have not selected the 'DB aggregation' check mark in Listcube.

Thanks,

Praveen