Technically, solving the asynchronous distributed consensus problem in bounded time is impossible. As proven by the Dijkstra Prize–winning FLP impossibility result [Fis85], no asynchronous distributed consensus algorithm can guarantee progress in the presence of an unreliable network.
