Redundant state machines in network elements

A mechanism for providing one or more passive state machines behaving in the same manner as an active state machine so that one of the passive machines will take over operation from the active machine if the active machine becomes inoperative. The active machine, in addition to responding to each excitation, sends the same excitation to a first passive machine, which accumulates the excitations and waits until the expiration of a first time T 1 to “pseudo” respond to the excitation. That is, the first passive machine responds to excitations as does the active machine, but its responses are discarded. Similarly, the first passive machine, upon each pseudo-execution, sends the same excitation to a second passive machine, which accumulates the excitations and waits until the expiration of a second time T 2 to pseudo respond to the excitation; and the second passive machine, upon each pseudo-execution, sends the same excitation to a third passive machine, which accumulates the excitations and waits until the expiration of a third time T 3 to pseudo respond to the excitation, and so on. When an active machine becomes inoperative, another mechanism replaces the active machine with a passive machine. All passive machines are in a circle and each is assigned a number that variously increases in one direction around the circle, e.g., clockwise. Each participant, a passive machine that may become an active machine, compares its number to the two numbers of the two neighbors, and if the number of that participant is smaller than both numbers of its neighbors, then that participant becomes active.