Consider a queuing system with two parallel servers and a single waiting line: customers join a single line if both servers are busy. They are eventually serviced by the first free server. If a customer arrives while both servers are idle, $($s$)$he then prefers server $1($who happens to work faster$).$

The inter$-$arrival time of customers are known to be distributed Uniform$(5,15)$ minutes.

The service time for the first server is Uniform$(15,20)$ minutes, and for the second server it is given as: