“…In [3], the bottleneck control problem has also been addressed in a simple Markovian environment, and the value of the maximum throughput rate from the system is obtained. In this paper, the bottleneck control problem has been restated in a non-Markovian environment, and the proposed solution algorithm is an extension and a generalisation of that given by [2][3][4]. In this paper, the assumption in [2,3] regarding the loading process, which was assumed to be Poisson, is relaxed, and the distribution of the interarrival time is assumed to be generally distributed.…”