Merge replication with identity value ranges using not for replication option

[Erdöl BIRAMEN]

I am using merge replication and want to partition the data using different
ranges for identitiy columns in combination with the NOT FOR REPLICATION
option.

I have enter these different ranges on the subscriber manually BEFORE I
register these servers as subscribers, as afterwards one is NOT allowed to
make schema changes on tables participating in a merge replication.

So I do all the necessary manual editing on the tables of my prospective
subscriber.

Then, when it comes to register this server as a subscriber, I am not sure
which of the following snaphsot options I should select:

1.) Drop and re-create the existing table
2.) Delete the data in the existing table
3.) Leave the table as it is

1.Question:
Wouldn't I loose all of my schema changes on the tables, if I choose 1?

2.Question:
Wouldn't I miss the merge replication triggers and guid-columns, which are
normally created during the course of the inital snapshot, if I choose 2 or
3?

Thanks in advance
Erdöl

[He Xiang]

Answer to Question 1: Yes, you would loose the identity range you entered
for that table.
to Question 2: a> guid-columns are generated at snapshot time,
not the first time when merge runs.
b> merge triggers and merge procedures
will still be there when you choose 2 and 3.
So it is safe to do so.

For your information, in SQL Server 7.5, you will no longer need to manually
set identity range for each
subscriber cause it is going to be automated in a smart way - a new range
would be assigned to a existing
subscriber after it is over booked due to large number of inserts.

thanks,
Xiang He
SQL Server Replication Team

Erdöl BIRAMEN <h875...@obelix.wu-wien.ac.at> wrote in message
news:uG$LbEuJ$GA....@cppssbbsa02.microsoft.com...