Ethernet Connectivity Fault Management (CFM)

How maintenance endpoints detect a broken Ethernet service even when access links remain up.

Maintenance domains and service endpoints

IEEE 802.1ag introduced Ethernet Connectivity Fault Management, now part of IEEE 802.1Q. A maintenance domain defines the scope of the monitoring. Within it, a maintenance association groups the maintenance endpoints (MEPs) for a service.

Maintenance intermediate points (MIPs) can participate in diagnostics along the path. Ethernet loopback verifies reachability to a particular maintenance point by requesting a reply. Linktrace can show the responding maintenance points along the forwarding path.

Detect a loss of continuity

MEPs send Continuity Check Messages (CCMs) periodically. A receiving MEP raises a defect when the expected remote CCMs stop arriving and the receive timer expires. Compare the expected peer, association, maintenance level and interval at both ends before interpreting a missing peer as a transport fault.

A break inside the provider network can interrupt the monitored service even while the customer-facing links report up. The MEPs check the path taken by service frames, which can span several links.

Try it in a lab

  1. On a CFM-capable platform, configure two MEPs in the same maintenance association. Establish a baseline with bidirectional test traffic and both devices receiving their peer’s CCMs.
  2. Interrupt the service between the two endpoints while leaving the customer-facing links up. Verify that test traffic fails and the affected MEP reports a loss of continuity after the receive timer expires.
  3. Restore the service. Check that CCMs arrive again, the defect clears and data traffic recovers.

Sources