http://www.perlmonks.org?node_id=985232


in reply to Sybase::CTlib vs DBD::Sybase

In my experience, the “overhead” if any that might be introduced by a database-API layer is negligible in all cases.   What matters is how you go about doing what you are doing.   (And here, the API-specific options do vary, as not all database implementations are the same.)

The biggest and most important rules of thumb that I can immediately think of are these: