We consider scheduling a shared server in a two-class, make-to-stock, close
d queueing network. We include server switching costs and lost sales costs
(equivalently, server starvation penalties) for lost jobs. If the switching
costs are zero, the optimal policy has a monotonic threshold type of switc
hing curve provided that the service times are identical. For completely sy
mmetric systems without set-ups, it is optimal to serve the longer queue. U
sing simple analytical models as approximations, we derive a heuristic sche
duling policy. Numerical results demonstrate the effectiveness of our heuri
stic, which is typically within 10% of optimal. We also develop and test a
heuristic policy for a model in which the shared resource is part of a seri
es network under a CONWIP release policy.