MDM Reference Data

[Former Member]

Hi all,

I have a question regarding reference data in MDM. The business scenario is that we are using MDM to consolidate Master Data for Materials across 3 SAP ERP systems. We are using standard Business Content delivered with MDM for out Repository, which includes all the customizing tables (that are standard in ERP) as flat table lookups.

The question I have is should we:

1) Populate the lookup tables will all the possible values for each of the 3 ERP system customizing tables.

2) Do we define a set of consolidated lookup values in MDM first for each ERP customizing table and then value map the legacy value to correct consolidated value on import.

E.G Material Group has Values “Mat1”, “MatGrp1”, and “Group1” across each of the 3 systems. Do we populate the material group lookup table with all these values?

or do we create on value "Mat Group 1" and value map the above 3 values to this new material group value?

How does our decision effect Syndication and/or Central Master Data Management (create and up in MDM Materials for ERP system)?

All help appreciated!

Regards

Con

[michael_theis · Accepted solution]

Hello Con,

If you have a look at the IDoc / XML messages that are send from one SAP system to another, you'll see that these messages do mainly consist of codes instead of speaking names. so if you plan the syndication / central data maintenance, you must have the correct, remote system dependent keys in MDM!

In your example if you would send Material Group "Mat1" to an ERP system that has code "MatGrp1" for this Material Group, the ALE inbound will fail because it is not able to assign the correct material group.

Regarding your possibilities:

1) Is always the easiest way to go. You can load the reference data for all remote systems into MDM. So you can be sure that you have all required keys in MDM. Nevertheless, if you do this, you should have a look at the look up tables after the data load. The reason is that you may have typos in the description of the values, thus you might have entries for "MaterialGroup" and "Material Group". When creating new records in MDM, this can be very confusing. So it makes sense to consolidate the data after the load.

2) You could do this, too. But in this case, you'd have to do a lot of value mapping during the import. And you have to know exactly which key information belongs to which look up value.

From my experience the first way is the one easier to go. You have the effort only once and you do not have to touch the object import afterwards.

Kind regards

Michael

[Former Member]

The syndicated xml files to the concerned Remote Systems should have the values of the Material group as recognized by that system.

If System A, has Material Group Mat Gp1, then the syndicated xml files should have MatGp1, irrespective of the value in MDM.

You have to turn on the Key Mapping for this table.

One way is to import the data as it is with the key mapping turned on.

In the Data Manager you will have 3 records. Each of the record will have one key mapping.

You can then use the Matching strategy to match and merge. While merging, the key mapping also get merged.

So in Syndication, depending on the Remote System name, the correct value is syndicated.

<b>Another approach is:</b>

Import (Add on the fly) one value which you want to preserve with Key mapping on.

Then do the value map of the other two values with the key mapping on for that table.