i have an identity column which is NOT part of primary key. i have multiple publisher one susbcriber transactional replication (all invoices from branches comes in one head office db). i have not specified different identity seed for each branch(publisher) so all of them start with 1 and in head office db i find duplicate values for them, but since its not part of primary key, i dont get any error. i now learn that we need to give different seed value for each of branch db, now my questions are 1. is there any problem if i dont specify different seed for each branch db, since they are not part of primary key. 2. can we change the seed of identity coulumn, even when we have some record already creaed in that table. 3. any other issues regarding identity column. further, i accidently deleted some records from a branch db, not i want to insert it back , i know how to use identity_insert to force identity column, but i want to know is there any problem using this approach, also when this is in same replication as informed above.