This section explains the holdoff timer feature for idle insertion on Ethernet controllers, describing its function, benefits, and use cases.
A holdoff-time is a fault management feature that
-
delays the propagation of local faults to the client when a trunk fault occurs,
-
inserts idles into the traffic stream from the trunk port to the client port for the configured duration, and
-
helps prevent unnecessary client port resets during transient trunk issues or PSM switchovers.
How holdoff time works
When a fault occurs on the trunk port, you can hold the propagation of local faults using the idle insertion feature. This feature is enabled on the ethernet controller by configuring the hold-off timer.
When the fault occurs on the trunk, idles are inserted in the traffic stream from the trunk port to the client port for the duration of the configured holdoff-time. If the trunk port remains faulty beyond the configured holdoff-time, a local fault is transmitted towards the client device. If the trunk recovers from the fault before the holdoff-time expires, traffic resumes.
This feature can be used on customer deployments to prevent reset of client ports during a PSM switchover.