We consider a discrete-time system comprising a first-come-first-served queue, a non-preemptive server, and a stationary non-work-conserving scheduler. New tasks arrive at the queue according to a Bernoulli process. At each instant, the server is either busy working on a task or is available, in which case the scheduler either assigns a new task to the server or allows it to remain available (to rest). In addition to the aforementioned availability state, we assume that the server has an integer-valued activity state. The activity state is non-decreasing during work periods, and is non-increasing otherwise. In a typical application of our framework, the server performance (understood as task completion probability) worsens as the activity state increases. In this article, we expand on stabilizability results recently obtained for the same framework to establish methods to design scheduling policies that not only stabilize the queue but also reduce the utilization rate, which is understood as the infinite-horizon time-averaged expected portion of time the server is working. This article has a main theorem leading to two main results: (i) Given an arrival rate, we describe a tractable method, using a finite-dimensional linear program (LP), to compute the infimum of all utilization rates achievable by stabilizing scheduling policies. (ii) We propose a tractable method, also based on finite-dimensional LPs, to obtain stabilizing scheduling policies that are arbitrarily close to the aforementioned infimum. We also establish structural and distributional convergence properties, which are used throughout the article, and are significant in their own right.
