Cisco Support Community
Community Member

Gigaethernet link Flap not obeying the configured keep-alive values

Hello experts

I have a 10G link in-between two devices configured with a keep-alive setting of 10sec (default value). I notice some flaps lasting for about 100ms are detecting thus the connection times out briefly. according to my understanding, the two interfaces at both ends should not see this flap since it is far much less than the 10sec keep alive setting. Unfortunately my customer connected using this link is able to detect the flap by his voip service going down. My question is, How am I supposed to configure the link/interfaces so as not to respond to the flaps before the configured keep-alive setting of 10 sec expires? Thus flaps of less than a sec should not be detected by the interface..

Everyone's tags (2)
Hall of Fame Super Blue

If you have a 10 Gbps fibre

If you have a 10 Gbps fibre optic link that "flaps" regularly, setting a keep-alive is the least of your worries.  


You need to check out and determine why the link is flapping.  They shouldn't.  It's fibre optic.  It's not meant to flap regularly.  

Community Member

hi Leo, thanks for reply. The

hi Leo, thanks for reply. The distance in between the two devices is around 300kms..thus fiber is provisioned by a fiber service provider. So ive got a worker and protection circuit on the transimission level. I did tests on switching from worker to protection and i experienced switch hits of about 100ms while switching from worker to protection..and aslo from protection back to worker. I experienced 2 ping timeouts during the switching times, which lasted for about 200ms, and noticed that the router interfaces both ends detected a "Rx link down" error. Why would they notice the link down..whereas the keepalive settings (10sec) is high above the switching times (100ms)..?

CreatePlease to create content