Author(s): Palmer J, Mitrani I
Abstract: We examine the optimization of a system where the servers in a cluster may be switched dynamically and preemptively from one kind of work to another. The demand consists of two job types joining separate queues, with different arrival and service characteristics, and also different relative importance represented by appropriate holding costs. The switching of a server from queue 1 to queue 2, or vice versa, incurs a cost which may be monetary or may involve a period of unavailability. The optimal switching policy in the case of bounded queues is obtained numerically by solving a dynamic programming equation. Two simple heuristic policies -- one static and one dynamic -- are evaluated by simulation and are compared to the optimal policy. The dynamic heuristic is shown to perform well over a range of parameters, including changes in demand.
Keywords: Optimal server allocation, Grid computing, Dynamic programming, Heuristic policies