ELSM monitors network health by exchanging various hello messages between two ELSM peers.
ELSM uses an open-ended protocol, which means that an ELSM-enabled port expects to send and receive hello messages from its peer. The Layer 2 connection between ports determines the peer connection. Peers can be either directly connected or separated by one or more hubs. If there is a direct connection between peers, they are considered neighbors.
If ELSM detects a failure, the ELSM-enabled port responds by blocking traffic on that port. For example, if a peer stops receiving messages from its peer, ELSM brings down that connection by blocking all incoming and outgoing data traffic on the port and notifying applications that the link is down.
In some situations, a software or hardware fault may prevent the CPU from transmitting or receiving packets, thereby leading to the sudden failure of the CPU. If the CPU is unable to process or send packets, ELSM isolates the connections to the faulty switch from the rest of the network. If the switch fabric sends packets during a CPU failure, the switch may appear healthy when it is not. For example, if hardware forwarding is active and software forwarding experiences a failure, traffic forwarding may continue. Such failures can trigger control protocols such as ESRP or EAPS to select different devices to resume forwarding. This recovery action, combined with the CPU failure, can lead to loops in a Layer 2 network.