Do you have a case where you have a named instance you’d like to migrate to a default instance? Is there pushback on doing so? Are there concerns that there’s too much work or that something critical might break?
Back when we tried to squeeze every drop of performance out of physical hardware, the announcement of named instances, meaning we could run more than one instance of SQL Server on a single physical server, was a big deal. When high availability meant fail-over clustering as the primary option, often we didn’t want idle physical servers, so we would install additional instances, which had to be named instances. Under optimal conditions, each physical server ran an instance and if we had an issue with one of the nodes, then its instance ran temporarily on another physical node.
In the current environment where we heavily leverage virtualization or even serverless technology, I don’t see many named instances get created unless they are legacy because reporting tools and integration points use that named instance name and there are challenges migrating from the named instance. Usually there is a question about how to migrate gradually because it’s too much work for a “big bang” and/or “we don’t know what we don’t know” and folks are afraid that there is a critical report, like one that pulls data together for the board, that we don’t want to take a change on breaking.
The good news is that we can set the named instance port statically and we can set that port to the SQL Server default port. We can then leverage the standard SQL Server client behavior when a named instance or port isn’t specified. I’ve written all this up in an article here: