have an db with about 15 tables that will handle many companies. no data overlap between companies. is it more efficient run-time to use one database and index each row by company id, and one database and partition each table by company id, or to create a database for each company?
it is a web-based app using persistent connections. no copying.