The term heartbeat refers to the vSphere HA mechanism that determines whether or not a host server is functioning normally. Each slave server in the cluster sends a heartbeat signal to the master server every second and likewise, the master server sends a heartbeat to each of the slave servers in the cluster every second. When the master node does not receive a heartbeat communication, it might mean that network communication has failed, but does not necessarily mean that the slave server has failed. To verify that the slave server is still functioning and able to support the running of virtual machines, the master checks its health status by using two different mechanisms:

A host is declared failed when the vSphere HA master node cannot communicate with it by means of a network heartbeat, or a storage heartbeat, and the host cannot be pinged.