Re: [Idr] Next Hop changes

In message <BAY17-F22CRl5gWpcRd0000af99@hotmail.com>
"john smith" writes:
>  
>  
> I am amazed no one has answered to this post.
>  
> I have never seen the changes in BGP causing changes in IGP, as we rarely 
> ever have cases where people do BGP-->IGP. It would bring down your network 
> though :)

What's the question you are looking to get answered?

> I also feel that the concept of things like HOT Potato come into play only 
> in the case when 1 protocol has to depend on the other. Though, if I look up 
> the defination of Hot Potato in some classic books like Data Networks by 
> Robert Gallager and Dimitri Bertsekas , it does not seem to be what is 
> referred to as "Hot Paotato" in some other papers and discussions that I 
> see. The case and examples in the books never talk about multiple protocols 
> to put forth the point of Hot Potato routing.
>  
> Perhaps someone could help clarify the definations in the 1st place,
>  
> -brgds
> JS

I have no idea who Gallager and Bertsekas are or what they wrote but I
think Sean Doran or Peter Lothberg originally referred to using the
IGP cost criteria in BGP MED as "hot potatoe" and passed across
providers as "cold potatoe" (what a legacy).  The definition should be
clear in draft-ietf-idr-bgp4-experience-protocol-04.txt in Section
7.1.1. "MEDs and Potatoes".

Curtis

ps - I like the paragraph:  

   Seemingly more intuitive references that fall outside the vegetable
   kingdom refer to cold potatoe routing as "best exit routing", and
   hot potatoe routing as "closest exit routing", though vegetable.

but it looks like the last two words don't belong there or were part
of something truncated like "though vegetable terminology was at least
initially much more common" (which we could probably drop anyway).

_______________________________________________
Idr mailing list
Idr@ietf.org
https://www1.ietf.org/mailman/listinfo/idr