Violation of Primary Key in Merge Replication

Question

I have one Publisher and one Subscriber. They have
a UK
(Col1, Col2), PK is col1 + col2, and
the MERGE
REPLICATION adds a ROWGUID. Sometimes
they insert the same row when they are offline ( they generate the same UK, it's not IDENTITY column), because
of it, they generate different ROWGUID for the same row. Then, when I run the JOB for Merge Replication, I have a PRIMARY
KEY violation. I need to use the UK instead of ROWGUID.
It is possible?
Example:
Table_1 (Publisher)
Name: John (UK)
ID: 1234 (UK)
ROWGUID: xxxxx (Created by Merge Replication)
Table_1 (Subscriber)
Name: John (UK)
ID: 1234 (UK)
ROWGUID: yyyyy (Created by Merge Replication)

Answer 1

Hi Alex R,
From your description, you want to use the UK(col1, col2) instead of ROWGUID, right? If so, it is not possible.
To use your own column as rowguid column in merge replication, your column should have the following properties:
• Should be of Uniqueidentifier datatype
• Should have the ROWGUIDCOL property set
• Should not allow NULLs
• Should have a default of newid()
Also please note that only one column can have the ROWGUIDCOL attribute in a table, in other words, we can only specify one column to be
rowguid column in merge replication.
Reference:
https://msdn.microsoft.com/en-us/library/ms152746.aspx?f=255&MSPPError=-2147217396
Thanks,
Lydia Zhang
Lydia Zhang
TechNet Community Support