DWDM transport networks are on the verge of supporting the dynamic set up and tear down of wavelength services. This is becoming possible mainly via node architectures with flexible add/drop port utilization, where a pre-deployed transponder can be tuned to any wavelength and transmitted towards any direction, and advanced management and control software, which provide the means to remotely reconfigure equipment end-to-end. In addition to quickly setting up the transponders located at the end nodes of the service, these capabilities are key to reduce service setup times in long-haul networks, provided a number of regenerators have been deployed in advance. This paper proposes a framework to characterize stochastically the fitness of each node as a placeholder for 3R regenerators in a DWDM long-haul network under dynamic traffic. Based on the network and traffic information available, different strategies are devised to determine the fitness value of each node. Network simulation is used to evaluate and compare the effectiveness of these strategies.