Page Comparison

Problem

The Preferred Publisher is ignored when I have it set up on record level consolidation (this would also apply to field level consolidation too).

I have 3 files from 3 sources. There are suplicates across these sources and I want to consolidate at record level based on e Preferred Publisher.

My Preferred Publisher's and their order are as follows:

• SystemA
• SystemB
• SYSTEMC

When I load data through the hub, the golden record is being created from SYSTEMC. Why is my Preferred Publisher order being ignored?

Solution

When you set up a Preferred Publisher in MDM, please note that Oracle creates the physical code (Publisher code) in upper case. Therefore, in this example the Publisher Code for SystemA is SYSTEMA, and for SystemB is SYSTEMB.

You must therefore make sure in your DI jobs, that you use upper case values for B_PUBID, otherwise the publisher is seen to be unknown. In this example SYSTEMC was in upper case so was accepted as the first valid publisher in the Preferred Publisher list (SystemA and SystemB were not recognised).

...

When loading data into the hub, a given publisher seems to be ignored in record-level or field-level consolidation in Convergence for MDM.
This publisher is used either in a Preferred Publisher or in Custom Ranking consolidation strategy.

Solution

The value stored and expected by Convergence for MDM to identify the publisher of a record is the Publisher Code, and not its Name. Best practice:

• Codes in UPPERCASE
• Names in MixedCase

There are several reasons for a preferred publisher to be ignored in a consolidation:

• The Data loaded in the SD_ table do not use the Publisher Code in the B_PUBID column. This publisher code is case-sensitive. The data integration job must use the case-sensitive Publisher Code. In most circumstances this should be ALL CAPS.
  If the value in B_PUBID does not match an existing publisher code, then this publisher is considered at the last position in the preferred publisher consolidation. Other publishers' values will always be taken into account for consolidation.
• When the PublisherID built-in attribute is used in a Custom Ranking consolidation, the literal values specified in the SemQL expression should be equal to the Publisher Code. These values are also case sensitive.
  If the SemQL clause does not contain a literal that exactly matches the publisher codes, the clause will not work as expected.

Related articles