[ceph-users] Questions about 'public network' and 'cluster nertwork'?

if the public network and cluster network use the same ip, why still need to send heartbeats to hb_front_server and hb _back_server at same time?