Skip to Main Content

Integration

Announcement

For appeals, questions and feedback about Oracle Forums, please email oracle-forums-moderators_us@oracle.com. Technical questions should be asked in the appropriate category. Thank you!

MDM: deduplication and handling old transaction data

e7544203-0f22-49a3-9c67-ab2a63d7582eApr 15 2018 — edited Apr 15 2018

Hi all,

I'm working for a financial institution on an MDM solution and have a question that seems basic but still I cant find an answer to it. Hope you can help.

The company has an IT landscape of roughly 20 legacy client administrations feeding into the new MDM hub and delivering client data downstream. In old setup, the transaction systems booked transactions directly against the legacy client ID's from the legacy systems. Now we are moving to a new situation with a unique global identifier. However, as we are deduplicating the duplicate client records _in_ certain legacy systems we find that the resulting client data with the global unique identifier becomes unusable for certain systems as they can no longer relate back to the legacy client id.

For example:

Suppose in the legacy systems we have duplicate client A that occurs three times. Then the MDM hub will deduplicate this to one golden cope and create a global unique ID and a mapping to source ID's like so:

Legacy client admin Client A                                MDM hub dedups client A                      Downstream transaction system

legacy ID: 1                                                          legacy ID: 1 global unique ID:1              global unique ID:1  transaction booked against legacy ID: 1             

legacy ID: 2                                                          legacy ID: 2 global unique ID:1              global unique ID:1  transaction booked against legacy ID: 2             

legacy ID: 3                                                          legacy ID: 3  global unique ID:1             global unique ID:1  transaction booked against legacy ID: 3             

However in the downstream systems that now receive the new global ID it becomes impossible to relate back to the original client data. This causes systems that have local copies of that client data to break.

My question what is the right approach to handle new global IDs propagating to downstream systems that depend on the old ID's, especially in combination with deduplication?

- Must data consumers always rebook transactions to the new unique identifier? Is this mandatory for a good functioning MDM solutions? Are there other ways / patterns to handle this?

- Is it ok to book a transaction against a unique customer identifier or should it be another identifier (for example an account)

Thank you,

Kind regards,

Hein

Comments
Locked Post
New comments cannot be posted to this locked post.
Post Details
Locked on May 13 2018
Added on Apr 15 2018
0 comments
233 views