This paper proposes APART (A Posteriori Active ReplicaTion),a novel active replication protocol specifically tailored formulti-tier data acquisition systems.Unlike existing active replication solutions, APART does not rely ona-priori coordination schemes determining a same schedule of eventsacross all the replicas, but it ensures replicas consistency by means ofan a-posteriori reconciliation phase. The latter istriggered only in case the replicated servers externalize theirstate by producing an output event towards a different tier.On one hand, this allows coping with non-deterministic replicas, unlikeexisting active replication approaches.%On the other hand, it allows attaining unparalled performance if replicas are often {\em silent}, i.e. if the average ratio between replicas' input and output messages is high.On the other hand, it allows attaining striking performance gainsin the case of silent replicated servers, which only sporadically, yetunpredictably, produce output events in response to the receipt of a(possibly large) volume of input messages. This is a common scenario indata acquisition systems, where sink processes, whichfilter and/or correlate incoming sensordata, produce output messages only if some application relevant event is detected.Further, the APART replica reconciliation scheme is extremely lightweight asit exploits the cross-tier communication pattern spontaneously induced bythe application logic to avoid explicit replicas coordination messages.