Currently cluster listeners behave the same way for replicated caches as distributed caches. We should enhance this so the replicated cache only uses a local listener to improve write performance when they are installed.