Hello,
Frequently I receive questions related to difference between standard parallel buffering and pseudo ascending in FI. So I decided to share with you some findings related this topic:
There are 2 differences.
1.a) Parallel buffering uses number of numbers in buffer as maintained in transaction SNRO.
1.b) Buffering in Pseudo ascending order is using always number in buffer = 1. This is hard coded for buffering in pseudo ascending order.
Usually parallel buffering with numbers in buffer = 1 solves performance / lock waits issue for RF_BELEG.
2.a) Parallel buffering: In case of rollback number(s) will be reused.
2.b) Buffering in Pseudo ascending order: In case of rollback number(s) will not be used again. Otherwise the number assignment would not be chronological anymore.
Numbers which were not used again are transferred to table NRIV_RESTE. So in case of Pseudo ascending order you may have gaps in BKPF. But these gaps are not real gaps as there is report RSSNR0S1 for documentation.
For an overall explanation related to Number Range in FI, I usually check the following WIKIs. They link the most relevant notes for each specific topic:
NUMBER RANGE: BUFFERS (Blog)
NUMBER RANGE : Buffering
NUMBER RANGE : Gaps in Number Range
NUMBER RANGE : General Information
NUMBER RANGE: DEBUG and CREATION
The below SAP note is a FAQ note and very useful also:
1398444 - Buffering the document number assignment for RF_BELEG
I hope this can help!
Regards,
Raquel
