At what point does our server say - Forget about it?

last week we had a big on sale and our server had issues.  Our api host asked us three questions which I am not sure how to answer.  It looks like the web harnessing tool would answer the questions, but I feel a little out of my league even reading the documentation.  So... wondered if someone can help out while I'm waiting for a response from Tess consulting... thanks.

How many concurrent connections is the Tessitura server able to handle?

What type of latency is created @ what number of connections?

What happens when the maximum number of connections is reached?