best way for handling identity columns in replication

Question

I need to implement Merge replication with a scenario like a Central Main server, 4 Office Servers located at each country office, and each machine having sql express insatalled on them.

Each server has SQL standard 2008 and Each client with sqlexpress will replicate data between assigned office server.

All OFfice servers will replicate the data amonghst themselves through the central server

What is the best way to handle Identity Columns?

60% of tables have identity columns.

I am worried that there will be a lot of conflict due to identity columns. What will be the best option??

Answer 1

See Replicating Identity Columns: http://msdn.microsoft.com/en-us/library/ms152543.aspx

Each replica will have it's own set of ranges for new items created locally.

Answer 2

My experience with merge replication is: much trouble. I would avoid that amount of separate sql servers if possible.

When you really have to replicate and merge data between sql servers i would suggest to use an uniqueidentifier column instead of an identity column. Merge replication needs a rowguid column anyway.