A disk server receives requests from many client machines and requires 8 milliseconds to respond to each request. Let N = the number of additional requests that arrive before the service interval is complete. The probability of exactly k additional requests in the 8-millisecond service interval is P(N = k) = e−0.9(0.9)k/k!, for k = 0,1,2,...,[infinity]. If 2 or more new calls arrive while the service interval is only partially complete, what is the probability that exactly 3 new calls will arrive before the server is ready to respond?
Hint: Use rules for conditional probability.