[c-nsp] BW parameter on the Interface

[Joe Abley]

On 20 Sep 2004, at 13:08, Jon Lewis wrote:

> On Mon, 20 Sep 2004, Rodney Dunn wrote:
>
>> The BW command doesn't have any influence over
>> packet forwarding in an equal cost paths scenario.
>
> Because it makes the costs unequal? :)
> I've seen BW (interface blah\nbandwidth $foo) alter how ciscos load 
> share
> across parallel T1s or choose which FE interface to use when there are
> multiple 'just as good as the others' looking choices.
>
> Perhaps in the cases I've seen it's because BW influences OSPF routing
> decisions?

If it does, it's news to me. I've never seen this.

I've seen comments before about how "interface loading" or "bandwidth" 
can influence path selection in EIGRP, but only in marketroid 
pseudo-technical documentation (hence my vague phrasing).


Joe