Design Principles for Flexible Systems

A fundamental aspect of designing queueing systems with stationary servers is identifying and improving the system bottlenecks. In this paper, the concept of a bottleneck is extended to queueing networks with heterogeneous, flexible servers. In contrast with a network with stationary servers, the bottlenecks are not a priori obvious, but can be determined by solving a number of linear programming problems. Unlike the stationary server case, we find that a bottleneck may span several queues. We then identify some characteristics of desirable flexibility structures. In particular, the chosen flexibility structure should not only achieve the maximal possible capacity (corresponding to full server flexibility), but should also have the feature that the entire network is the (unique) system bottleneck. The reason is that it is then possible to shift capacity between arbitrary points in the network, allowing the network to cope with demand fluctuations. Finally, we discuss how knowledge of the system bottleneck may be used to decide how to add server flexibility to an existing network.