Etherchannel Misconfig guard

[xZamalek]

Dears,

Lets suppose that S1 is hard coded to be an etherchannel and S2 doesn't have etherchannel config , how the loop will occur?

in case that:-

1-A device connected to S1 send a broadcast or unknown unicast , will the etherchannel follow the default load balancing algorithm and choose only 1 link inside the Port-channel to forward the frame? or in this case will use both?

2-once the broadcast frame goes to root , it will flood the frame out of all ports expect the port it received the frame , S2 will send back the frame from the other port to S1 , however S1 will detect it received the frame on "Po1" as a whole not fa0/1 or fa0/2 so its impossible to transfer the broadcast back to s2 , so how the loop will occur in case of etherchannel misconfig?

[luis_cordova]

Hi @xZamalek 

Check if this discussion can give some help:

https://community.cisco.com/t5/switching/lacp-vs-etherchannel-misconfiguration-guard/td-p/2994843

Regards

[xZamalek]

Thanks a lot , i will check this discussion

[Jaderson Pessoa]

well, in shot awnser. all ports will be blocked, because from one side has LAG enabled and from another side havent it.

1-A device connected to S1 send a broadcast or unknown unicast , will the etherchannel follow the default load balancing algorithm and choose only 1 link inside the Port-channel to forward the frame? or in this case will use both?

R: in your exemple, switch s1 will disable po1 because negociation between switches failed, both ports will be disabled.

[xZamalek]

will it block the Port channel even if the etherchannel was hardcoded? as far as i know hardcoded port channels doesn't generate negotiation messages , maybe i am wrong.

[Jaderson Pessoa]

Well.. what you want to say about hardcoded?