Characteristic without valid master data still being loaded

[Former Member]

Hi experts.

I have a process chain which loads an ODS with data. Among other data I have 0COSTCENTER. The load process is through PSA and after that to destination by packages. I have the "do not update data when no master data exists" option selected.

I have usually a few records with characteristics that do not exist in master data marked as errors which are not updated to the destination ODS. That is expected and intended.

But recently someone posted an erroneous document in the ERP where they filled mistakenly the Cost Center. When I extracted this document it tried to fill 0COSTCENTER with the value "369.123,27". Since this value is not in the master data table of 0COSTCENTER I would expect the load process to mark this line as an error and keep it in the PSA as such. It should not try to update the data target with this line!!!

What is really happening is that the PSA error lines do not contain this line. Instead it has been transfered to the data target but there was an error during the activation stating that the value is not valid for 0COSTCENTER.

What is happening? Shouldn't this line be automatically marked as an error line and not be transfered to data target? What can I do to prevent this (obviously without depending on ERP users to fill in correct data all the time...)

Thank you. Points will be awarded for helpful answers.

[Former Member · Accepted solution]

See if Bhanu's suggestion at will help you.

Regards,

Gili

[Former Member]

OK. I'll give a try. Tomorrow I'll have results since our process chains run daily at night.

[Former Member]

I've read Bhanu's post on the thread you've indicated.

"When using Ref Int check, be sure to use Only PSA and Update Subsequently in Data Targets as the processing option, and turn on Error Handling in the Update tab to Valid Records Update, reporting possible (request green). "

I confirm that I'm using Referential Integrity and error handling with reporting possible. I'm not using Only PSA and Updade subsequently but instead I'm using PSA and then into data targets (package by package). I do not think that this difference is relevant for this matter. Please correct me if I'm wrong.

Then thing is I do have some entries which are marked as error and I can correct them in PSA. This is already working and it didn't work for this record!

Another strange thing that maybe has nothing to do with this but I should relate anyway. The R/3 people here cannot replicate the erroneous document. According to them the system has a standard check which should not allow wrong cost centers to be posted. They could not replicate the wrong document in the quality system. Very strange indeed.

Any other ideas?

[former_member188975]

Hi Pedro,

"I'm not using Only PSA and Updade subsequently but instead I'm using PSA and then into data targets (package by package). I do not think that this difference is relevant for this matter"...Actually it is relevant and you do need to use the thrid option in the Processing tab of the InfoPackage, if you need to separate the error records and allow the green ones to pass through.

Hope this helps...

[former_member188975]

Hi Pedro,

Check for the following in your InfoPackage: Processign tab has PSA and Subsequent update option and the Update tab has settings for Error Handling. Only when you set error handling (in your case use the reporting possible options) will the incorrect records be collected separately in the PSA.

Hope this helps...