Calculate database performance on overload

I don't know if this is the right place for this question, but here it comes:

I have a 开发者_运维问答db function that I tested and can handle about 3000 requests/minute. The problem is that this calculation just gives me the optimal performance of the function (~18ms/request). How do I calculate the performance when the requests are twice, triple or n-times as that?

try creating multiple threads or processes that all call the function as fast as they can. preferably use the same number of threads that would be used in a production environment.