I have a published database installed with about 60 users.

The update process is one-day, so Users send Update files to the server and they are loaded there, no data comes back in the opposite direction.

There is a need to add a new table to the central server which receives data from all users. This table will not be present on the Users' databases.

My question is: Will the Updates from the Users' databases load successfully, or will the existence of the new data-tables make the process fail.

My understanding is that the Updates process works atomically by data row, so provided that each row update CAN apply to the database, it will. However I am not 100% certain, so a bit of clarification would be great.

> My question is: Will the Updates from the Users' databases load successfully, or will the existence of the new data-tables make the process fail.

They should load without any problems.  With my replication setup, I can add a new unpublished table into either end of the replication process without it complaining.