When I started working on replication several years ago, I was convinced that identity range management was only available on the latest SQL servers for legacy purposes, since before SQL Server 2000 there was no GUID type field and most primary keys were built on auto-increment fields.
I have already seen that some of these databases "work", and I can say that managing these IDs in a complex replication topology can be a real PITA, guaranteed by headache and debugging sessions throughout the night.
I just read here that Identity Range Management was still offered by default in SQL Server 2008 , followed by this other question about best practices and limit range management , where # mwolfe02 concludes that in some situations, users cannot insert data to the database, unless they have been granted db!
Anxiety, isn't it. So, with the exception of legacy databases, why should someone use identity range management with SQL servers? More generally, why does someone advertise some predefined method for identifying auto-increment in a database?
source
share