Skip to Content
0

Processing timesheet records taking longer time to process

Mar 16, 2017 at 09:43 AM

24

avatar image

Hi Experts,

We are working on PI 7.3 dual stack. The requirement is all the employees enter the time sheet details in an app of organization and PI pulling the data from SQL view via JDBC channel and posting the data to SAP S4HANA by ABAP proxy. The challenge here is say atleast 40k employee are filling timesheet on weekly basis[say on Friday] which creates around 40k*5 equals to 2 Lakh records on that day. In order to process that large number of records in short time what best can be done from PI perspective? What better can be done in S4HANA (or) ABAP team perspective to reduce the processing time?

Thanks,

Ramu G.

10 |10000 characters needed characters left characters exceeded
* Please Login or Register to Answer, Follow or Comment.

1 Answer

Andrzej Filusz Mar 16, 2017 at 10:40 AM
0

Hi Ramu,

I recommend an asynchronous scenario. On the SAP side I would create a table ("staging area"). In the proxy, all data from a message would be inserted into that table without any processing. Then I would write a report, that would work in the background and take unprocessed rows from our table and process them. After a successful processing of a row I would change its status. To reduce processing time, a report could work in parallel (PERNR ranges for example). If your data must be inserted into infotypes, keep in mind that before making any changes for a employee, you must lock that infotype record and then release it. Sometimes it can't be done because, for example, a HR worker can modified this PERNR manually. So you must think about reprocessing those rows for which setting a lock failed.

Regards,

Andrzej

Share
10 |10000 characters needed characters left characters exceeded