This paper addresses the consistency and scalability behavior of a soft-state signaling protocol. Soft-state is state that has a predefined validity period. A soft-state protocol is a signaling approach in which soft-states are periodically refreshed by the receipt of a message indicating a new validity period and possibly a new value. Unrefreshed state eventually expires and recovers as soon as a subsequent refresh message is received. This protocol is widely used in distributed systems where components need to keep track of the state of other components. Within such a system, the state consistency is affected by many parameters, some given by the environment while others are tunable. The environment parameters are the loss probability and latency of the channel between the endpoints, and the change rate of the source state. The tunable parameters are the refresh period and timeout period. In this study, we dissect the overall system inconsistency into false-positive and false-negative cases, address the tradeoff between the two, and analyze their sensitivity towards parameter changes. In the end, we identify the necessary measures to optimize each of them. Our results show that false-positive ratio can be optimized by using a state cancelation mechanism, whilst the false-negative ratio can be optimized by specifying the timeout period as a function of the other parameters. Moreover, we study the system’s scalability behavior by observing the growth of messages overhead that is needed to keep inconsistency ratio constant as the channel’s loss probability increases.