Handling IDENTITY column in Merge Replication

[Rajeev Pandey]

Hi,
I am having a table containing an IDENTITY column as Primary Key(PK)
and another table referring to this PK. Now when i am trying to
replicate through Merge Replication, it mis-mathches the data, i.e., the
record which i have inserted against one PK, after replication it is
replicated against another master record.
If any one having idea about handling these type of IDENTITY columns
then do let me know
Thanks
Rajeev

[Edmund Shaw]

Use NOT FOR REPLICATION to stop ids being reset.
Use different id ranges (or a composite PK eg with a servername field) at
different servers if inserts are done in multiple locations to prevent PK
violations.

Actually, BOL is good on this. Get used to using the search tab and you'll
go far...

Edmund.

Rajeev Pandey <pandey...@hotmail.com> wrote in message
news:38F675F48B36D4118E4D...@cpmsftmsgv21.microsoft.com...

[Ryan Stonecipher]

Actually SQL 2000 has excellent support for this. Will dynamically assign
ranges to IDENTITY values on the Publisher and Subscriber to handle this.
Another idea is to assign your primary key as a column other than the
IDENTITY column, perhaps a ROWGUID column.
Ryan Stonecipher
Microsoft PSS, Developer Support Database
ryan...@microsoft.com