Excel MD Mass Maintenance restriction...

[former_member222291]

Hello Expert,

As probably many of you, our IBP model requires that some master data be maintained natively in IBP.

During testing activities, I noticed that the Excel Mass Maintenance tool was limited to a certain number of records.

I configured MAX_BATCH_SIZE global parameter to 900 000, but as soon as I try to load more than approximatively 290 000 entries, I receive an IBP error message : "An error has occured on the server". Here is the detailed log:

     2016-02-19 12:42:38,717|ERROR|UILayer|?.?|xxxxxxxx|||||VSTA_Main|From SOP - Backend / API Error: An error has occured on the server#

     2016-02-19 12:42:38,758|ERROR|UILayer|?.?|xxxxxxxx|||||VSTA_Main|From SOP - Backend Error: An error has occured on the server#

Since we are planning to maintain Customer Sourcing Ratio with this tool, this is kind of a issue. We have around 4 millions possible sourcing combiniations. And even when using attributes filter, we far exceed the limit. We would need to be able to display at least 650 000 records.

I understand that Excel is limited to +- 1 000 000 lines, but is there any additional hard coded limitation in IBP engine? I am aware of the following warning from SAP, but we don't aim to replace standard integration... just using it as a complementary tool for our MD maintenance.

Any advise/experience sharing is welcome.

Thanks and regards.

[Irmi_Kuntze · Accepted solution]

Dear Philippe

No I have not heared of such limits.

Are you on IBP 4 or 5?

I cannot give you any advise, I am just curious about the business case behind:

How do you get the MD into the system for the initial fill, by webui, manual work, HCI load from somewhere?

Do you "just" want to load all customer ratio into the tool, or maintain them all at the same time?

600'000 is quite an unsusual requirement for manual maintenance, and for a GoLive I would anyhow rather upload than maintain

If I where responsible to maintain such big numbers, I would want to do that in smaller pieces anyhow as I would not be able to do 290000 entries at the same time, as the danger that Excel crashes and all my work would be gone would be too big for me to take.

If you have that as a hard constraint, and an OSS message would not help, than in your case I would "invent" some extra attribute in the customer master that allows to bring it to smaller pieces. If the customer's country does not bring it down in smaller pieces, and if combining that with some customer group does not help either, an additional group for seperating may help.

Yours Irmi

[former_member222291 · Accepted solution]

Hello Irmhild,

We are running on IBP 5.0 SP6.

Our customer required some flexibly in order to be able to plan all imaginable Customer-Product combinations, i.e. every customers could buy every products...

In addition, they also required to plan their supply based on preferred sourcing location, but with the possibility to adjust manually the operator result.

Again, they ask for flexibility, so that they can choose supply source among all plants or DCs where stock is available.

This results in 1 million+ Customer-Product combinations and nearly 4 millions Source Customer entries.

Our initial plan was to generate every possible Customer-Product entries via HCI, and then, generate Source Customer Master Data by joining ECC-MARC with the previous table.

However preferred sourcing location is not maintained in our customer's ERP, that's why we need an hybrid data maintenance.

After some discussion with the business, I changed my mind on this and I think I found a more viable way to proceed:

     1- We will remove Customer-Product from our planning level, since in our context, we don't need to      manage any restriction. Like this, we will avoid to maintain useless data in IBP.

     2- Only historical Source-Customer relation will be loaded initially in the system. Then, Supply      Planner will have to run a Check Mode operator at the beginning of his supply planning cycle in      order to capture missing data. He will only have to maintain preferred sourcing location and we will      generate with HCI magic the alternative sourcing location.

I was also considering adding some extra attributes to segretate our data more easily.

To your knowledge, could Virtual Master Data be an option? I noticed that some attributes can not be use as filter for Mass Maintenance... it seems to be limited to attributes that are assigned to planning areas only ...

Regards

Philippe

[Irmi_Kuntze · Accepted solution]

Dear Philippe

Glad you thought about another approach, sizing would need to be considered as well ....

Yes, if you want to seggregate data by additional attribute, you have to include that into the planning area as well, not only the MD type

If it is only in MD type, I observed that it will show up in MD maintenance in Excel, but you cannot filter for MD maintenance nor use it as selection criteria in PLanning View itself.

And the attribute must be in minimum one planning level, even if no KF is assigned to that level.

Virtual master data would work as follows:

If you for example have a customer grouping on your single customer, and you want to have the same sourcing per each customer within the same group, than by working with virtual MD type you can maintain attributes on aggregated customergroup - source - level, and the system would automatically move the attribute to customer - source - level

This can help reduce the entries a lot and solve the issue with the millions of entries.

But it can not handle exception if e.g. one customer within the group should get a different attribute entry

And here again, you would need to find some kind of reasonable grouping

Whatever way you go, thinking about customer grouping is a good idea...

Yours

Irmi