The best way to harmonize !!!!

[Former Member]

We are trying to harmonize a scenario between a crm system with MDM : After processing the data with data manager , we want to syndicate the data .What is the best way to process ?

1- Send the data to the same CRM SYSTEM. ? if yes does it make sense to process and update the data of the same system at the same time ?

2- I have to client systems. I send the data from client system 1 to MDM. The client system 1 has a table called Bussiness Partner and a ID like GUI12345. The second client system has its data of Bussiness Partner with a ID like IDD4567. How does MDM manage the key transformation from client system 1 to client system to in an harmonization scenario?

3. Now whe have 2 clients systems again. The client system 1 has a ID like 555 and the MDM sends that data to client system 2 that also follows this ID type. If in the client system 2 the data with ID = 555 already exists.. what happens? How MDM manage this situation?

Thanks in advance,

Regards !!

[michael_theis · Accepted solution]

Hi Stan,

1) This depends on your scenario. Of course you can load data from CRM to MDM, change / correct / add data in MDM and the update the same CRM with the revised data.

2) The answer to this question is "Key Mapping". If you import data from system 1, MDM is able to store the key for this record. So if you send the data back to the same system, MDM uses the available key to update the existing record. Regarding client system 2, the question is now, if the record already exists in this system, too. If yes, you can add the according key mapping information. If not, follow Annika's description to define a number range that is used by syndicator to create new keys.

3) MDM will update the existing record. This will happen due to the default behaviour of the remote system inbound processing (ALE or Proxy). In harmonization or central data management scenarios, the careful and correct definition of the number ranges is the most important task to fulfill!

Kind regards

Michael

[Former Member · Accepted solution]

Hello all,

According to this question, I understand that first to all I have to define a Range in my console.

Imagine that client system one works with the range (from 1 to 1000) and client system two works with the range (from 1 to 2000).

Actually client system one has its loaded data from 1 to 50 (so, from 51 to 1000 is free space) and client system two has data from 1 to 70 (so, from 71 to 2000 is free space).

In the console, I should stablish for the client system one range from 200 to 1000 (the new records that MDM syndicator sends to this client system will be saved in the 200, 201, 202...etc position).

For the system client two if I stablish the range from 500 to 2000 the new data that comes from CRM will be saved in this defined range.

If a data already exists in the client system one, it will be uploaded (from 1 to 50) and for the client system two (from 1 to 70). It should be declared in the key mapping tool.

Is this correct?

Finally, my new questions are:

1. What happens if I don't declare a range? MDM begins to save date in 1 position or is the client system wich generates the new key?

2. What is exactly the process to add a key mapping (from the client system twoo)?

Thanks in advance,

Marta

[Former Member · Accepted solution]

Somethings is not clear in my mind :

Imagine that I am working with MDM and two systems (1 -CRM 2-ERP)

If I understand well : First of all I have to know the range which was established in the systems CRM (i.e 0.......10000) and ERP (0.......5000 ) .These ranges are the same that I have to define in my console repository when I am defining my remote systems CRM (0......10000 ) and ERP (0........5000) ?

In the case that CRM is using the range 0 ... 50 (records) and ERP uses 0 ...70 . When MDM is harmonizing if the records already exist in the system 2,we can add the according key mapping (i.e record 0...50 will be updated ) and if not we have to create new records ( 51..52..etc...) and the same scenario will occur for ERP (71,72...etc )

I am right ?

[michael_theis · Accepted solution]

Hi Marta,

1) If you do not declare a range, the sysndicator won't set an automatic key. You record will be send to your remote system without keys. Then it depends on the remote system if the data will be posted or if the data will fail. Each SAP system supports two different kinds of number ranges. One has "internal number assignment" which means that the system will define a key itself (then MDM must not send a key!), the other has "external number assignment" which means that MDM has to provide a key.

2) There are different possibilities. You can always add key mapping manually using the Data Manager if the records already exist in MDM. You can use the key range together with the Syndicator as well. Then the syndicator would create a key during the syndication. If the records do not exist in MDM, key mapping has to be written during the import of the records.

Kind regards

Michael

[michael_theis · Accepted solution]

Hi Stan,

if already records exist in CRM and/or ERP you should load those as a first step to MDM. Then you'd have the already used keys in your MDM system. Then you should define the number ranges according to CRM and ERP. This is correct. If you now syndicate the data back to CRM and ERP, the existing records would be updated (since the key mapping is already in MDM), and new records would be created with the next free key from the corresponding key mapping range.

So your assumption is correct.

Kind regards

Michael

[Former Member · Accepted solution]

Thank you Michael,

1) My client system has a internal number assignment, so I understand that If I don't define a range and a new data comes to the client system, the client system will generate the new client system key automatically :-).

2) About the key mapping generation... I am working with the next Harmonization scenario:

1. A user changes a date (with ID = 35) in client system one.

2. All data are imported into MDM with use of import manager.

3. I look at the new data and consolidates it with another data ( with ID = 55) that belongs to the client system 2 and that is a duplicate.

4. I use key mapping and stablish that the new key default will be 55 (corresponding to the ID of the client system 2)

55 client system1 35

55 client system2 55

5. I syndicate the new consolidated data to the client system 2 ( so, the 55 position will be updated).

a) I suppose that in the syndicator manager I only have to syndicate the new consolidated data to client system 2 (I have to make a match between the generated key and the ID of the client system). Is it correct?

b) I suppose that I only can syndicate data to one client system (that correspond to the key mapping generation). Is it correct?

c) How can MDM syndicate not consolidated data? i.e. client system1 sends a new data that doesn't correspond to any data loaded in the repository and I want to send it to client system 2 in a harmonization scenario...

In that case a key mapping is not possible... so I suppose that I don't have to map the ID in the MDM syndication. By this way, client system 2 will interpret it as a new data and will load it automatically in its new position. Is it correct?

Thanks again!!

Marta.

[michael_theis · Accepted solution]

Hi Marta,

all your assumtions are correct.

To clarify: key mapping consists always of two values:

- first the remote system ID

- second the record key for this remote system

An example:

You have one remote system with ID = RS1 and a second one with ID = RS2. You import one record from each remote system, record key from RS1 = 35, from RS2 = 55. In MDM you have both records, let's say with MDM ID 1 (for the RS1 record) and 2 (for the RS2 record). Each record has only one key mapping. Using the matching you find out, that those records are duplicates. You match and merge the records. this results in MDM into a new record with MDM ID = 3, that has now two key mappings, one pointing to RS1 with key 35 and a second one pointing to RS2 with key 55. If you now syndicate the data and choose to send the data back to RS1, syndicator will use the key mapping for RS1. If you do the same for RS2, syndicator will of course use the key for RS2.

If you run a syndication, you always have to define a target system. This can be only one system. So if you wan tto syndicate to two systems, you need to syndication maps and you have to run the syndication twice.

If you import data from RS1 with key 35 and want to send it to RS2 but this would be a new record in RS2, you may have to add the keymapping (manual or with the number range in syndication) or you can send the data without any key - which depends if you have to provide a key, or now (as explained before, this depends on the internal vs. external number range assignment).

Kind regards

Michael

[Former Member · Accepted solution]

Thank you Michael!Your answer has been ver helpful!

[Former Member · Accepted solution]

Hello Michael !!!

Many Thanks !!!

Wonderful you are a CRACK !!!

[former_member192347 · Accepted solution]

I am surprised, to read that Merging records in Data Manager creates new record with new MDM ID. Is there any configuration/process to setup that will generate new record with new MDM ID?

Also after merging two records from two different systems, when try to Syndicate them (though both Remote systems are selected) syndicater only generates only one output record for the default Key System. Does the SYndicator allow to syndicate two merged records from two different system at once?

How are you applying the Syndicated records to CRM (CRM 5.0)? SAP delivered content, CRM API, custom code or something else? Are you using XI as middleware?

Also have issues on extrating Business Partner data, inculding the Marketing Attributes?

Any experiences you want to share?

Since I am aslo working on scenario to integrate CRM to MDM, and there are lot issues around it. It's time consuming/diffcult to type all details here, can we setup a conference call and share our experiences.

Send availability, contact # and time zone

Phone:973-630-1535

Email: abhay_mhatre@colpal.com

Time Zone: US EST

Thanks!

Abhay

[Former Member]

Hi Stan ,

Usually in Harmonisation , Diff number range is used of source systems which is not used in any system .

Like sys 1 using 00000 to 55555

sys 2 using 00000 to 44444

Than in harmonisation use 60000 to 99999 so both sys have same data .

Hope this helps a little ,

- An