Inventory load with LSMW and MBGMCR

[former_member291014]

Colleagues,

We are using LSMW with idoc MBGMCR to load inventories for go-live and it works fine.

The issue is since we have alot of of inventories to be loaded we would like to create multiple sessions or start idocs in sequence and and the parallel process.

Our logic would be to seperate the files by material number.

The problem is that everything works but we do not get parallel processing, everything goes one after each other.

Our developpers are telling us that it is not possible to post multiple MBGMCR at the same time.

Is this true ? how can we increase our load speeds.

Thanks in advance

Eric

[former_member291014]

In our case separating by sku number is very easy. Biggest plant only has 15 sku's but wer are running batch management with characterisitcs had we have 50,000 of them.

The step ''BATMAS'' to create and classify is very fast, our issue is MBGMCR which takes 12 hours. The objective was simple has let's  split in 2 one file with 7 materials and the other one with 8.

Issue is SAP will not process 2 different materials in the same plant (or even different) plants at the same time... That is according to my dev guys.

[JL23]

As said one and the same  material in 2 IDOCs in parallel will create locking issues, but 2 different materials in 2 idocs in parallel have nothing to do with each other, and if you have enough processes then they can certainly post at the "same" time.

Here is a SAP note that explain in general the same: 924164 - Parallel posting of goods issue from the DC

[JL23]

From my point of view is separating files by material number the difficult step, is this really necessary? Why has it to be by material number? Can't it just be packages à 5000 as example?

You cannot post multiple IDOCs with the same material number at the same time. But will it really be the same? One posting process will wait until the other has finished if you really have 2 inventory figures for the same material.

If you have 50000 records in the source file, then you can run the conversion step 10 times, e.g. from 1 - 5000, from 5001 to 10000, from 10001 to 15000 ....

of course you need to create Idocs from the converted data before you convert the next part.

You can as well convert all in one shot, and can then execute IDOC processing in parts and schedule its posting in parallel in background.