Skip to Content

Hana optimized inventory infocube - implications and benefits


Regarding the implications and benefits of converting an inventory cube to a Hana Optimized cube:

I have read the following document:

Which really does a good job of explaining the various implications for standard procedures involving the optimized inventory cube.

What I understand from this document is that the the "historical" record type (2) represents movements that are older than the initialization records, while the "deltas" record type (0) represents movements that are newer than the initialization records. This is made clear by the explanation in page 16, where it is understood that the calculation of the stock quantity treats records in those two types differently.

The question that comes to mind that isn't covered in the document is whether the consistency of that representation is kept by the standard procedures, and more specifically:

1) Assuming the marker in a non-converted cube is updated so that all deltas are "older than" the marker, will all the movements move automatically to the historical record type when converting the cube?

2) When reloading the initialization record type (1) from 2lis_03_bx, will the relevant movements "move" from record type 0 to record type 2, or does one need to reload those movements as historical movements?

3) Does loading historical movements from 2lis_03_bf affect the record type of existing deltas, or does one need to delete existing deltas and re-initialize the marker?

Some other questions regarding this change:

4) Other than faster loading to the cube and easier remodeling, is there any specific benefit to converting inventory cubes to HANA optimized cubes?

5) As the change to HANA-optimized essentially means "current day stock" calculations now take movements into consideration, as opposed to a (HANA non-optimized) state where the marker is always updated to the current day, will converting the cube impact the performance of current day stock calculations?

6) Currently, in order to allow fast reporting, we materialize stock conversions to other units into key figures during loading. This has the negative impact of creating inconsistent and incorrect data when the conversion ratio is changed. Can on-the-fly conversions in the query be pushed down to HANA? What are the limitations or prerequisites of this (e.g. query properties that prevent that kind of pushdown). Is there any difference between non-optimized and optimized cubes in that regard?

Thanks for any help,


Add comment
10|10000 characters needed characters exceeded

  • Get RSS Feed

0 Answers