I was recently asked to take a look at at some SQL Servers for a friend and carry out a few config and security audits.
There were a couple of servers belonging to the same application group which interested me because there were many inconsistencies and generally odd configuration options set.
One thing which attracted my interest: Named instances configured on a dynamic default port (i.e tcp port 1433). SQL Browser was running but clients connected using ,.
My thoughts on this were that this was totally wrong. I mean, why bother creating a named instance only to put it on a default port, and why make that port dynamic?
I contacted the vendor who very assertively told me that their documentation and installation guide stated that instances should be named and configured on dynamic port 1433 ("As you know, this is the default port that connections expect SQL Server to be listening on and changing this would void your client's support contract with us...")
I can't bring myself to fashion a reply at the moment. Part of me feels it's not worth the argument as I'm doing this for a favour really.
To arrive at the point, my question is two-fold:
When would it ever be useful to operate named instances on default ports, and why would it ever make sense for them to be dynamic?