[c-nsp] MultiChassis LACP

[Phil Mayers]

On 07/10/2011 02:52 PM, scott owens wrote:

> If it is two switches / two chassis / two boxes connected in some fashion
> with fiber, copper, magic switch cable and it can provide a lacp link that
> can survive a failure of one of the devices ... it is a multi-chassis link
> aggregation capable device as far as I can figure.
>
> I have 7010s, 5020/5010s, 5548s, VSS, 3750s and they all look just the same
> to the end device - channel-group xyz mode active.

Terminating a LACP on >1 unit in a 3750 stack is certainly useful, and 
has similarities to multi-chassis LACP. I don't think anyone is 
suggesting otherwise.

My original email was in response to some fairly clear confusion where 
people were asking "can the 3750 really do this", and was intended to 
clarify that it depended on what you meant by "this".

I think it proves my point when someone else chimed in "I agree", and 
then you replied saying "I don't". There is disagreement about the 
terminology, and it's therefore best to be specific, as "it's all the 
same" for one person is another person's "it's completely different".