“…Thus their model is a combination of the M/G/1 and M/D/1 queues and the server keeps switching over these two queues depending on the class of units present in the system. For separate references on M/G/1 and M/D/1 queues, the reader is referred to Bhat [10], Levy and Yechiali [11], Kleinrock [12], Cohen [13], Lee [14], Gross and Harris [5], Cox and Miller [16], Tijms [17], Yang and Li [18], Bunday [19] and Madan [20,21]. However, in the present paper, we generalize Madan and Abu-Dayyeh [9] paper by adding a significant assumption to their model that the server may take a vacation of random length but we assume that no vacation is allowed if there is even a single priority unit present in the system.…”