ANNOUNCEMENT: NANOG 9 Date Change (fwd)
R. Eric Bennett
reb at ieng.com
Tue Nov 26 15:36:06 UTC 1996
> At 9:37 AM 11/26/96, Avi Freedman wrote:
>
> > Route reflecting sounds like a good topic - could I interest any of you
> > in presenting on it?
>
> > ------------------------------------------------------------------------
> > Susan R. Harris, Ph.D. Merit Network, Inc. srh at merit.edu
>
> I would be willing to present, though as I said I think a separate meeting
> to see what people really want is needed.
>
> I think the issues are:
>
> o (Briefly) The politics and technology of peering
> o Easier peering between multiple parties: MLPA
> o Since no NAP operator is going to enforce an MLPA, how can peering between
> multiple willing parties still be made to happen with less people time
> involved in the setup?
> o Why might the RA not be the best tool - or why might it be?
> o Possible goal:
> o Participants sign a contract expressing a desire to peer with anyone
> else signing the contract (not exclusively) through a route-reflecting
> box.
> o You can only offer routes for you and "your customers" via this. No
> partial transit to specific people can be offered.
> o Boxes at each interesting exchange point that people can then peer with
> to effect the agreement. One or two Cisco 2501s would work fine, but
> RA-type boxes which can "hide" their ASs in the middle might be
> interesting as well (Peter Lothberg arguments about BGP not being
> designed to 'work that way' possibly put aside).
> o Filtering:
> o Box-side filtering to enforce sanity?
> o Concerns
> o Who's going to run the thing?
> o Network stability?
> o What happens to control bad neighbors?
>
> Or, perhaps a separate mailing list is needed in the interim to allow
> people to discuss the issue without boring uninterested members of
> the nanog list...
While your outline sounds great wrt its chosen topic, the topic doesn't
sound like what I consider to be route-reflecting -- specifically, route
reflection in (i)BGP. Your outline sounds more like "politics and
operational issues surrounding peering and route-serving at a NAP." Can
someone clarify which of the two topics is the burning topic that people
would like presented?
Note that both topics may be burning issues and worthy of a presentation at
the next NANOG...
thanks,
eric
----
R. Eric Bennett <reb at ieng.com> | Internet Engineering Group
313-669-8800 (v) 313-669-8661 (f) | 122 S. Main, Suite 280
http://www.ieng.com/ | Ann Arbor, MI 48104
"Radical Rodent: Superdynamic Rodent of Tomorrow"
-- http://home.earthlink.net/~krhughes/Rat.html
More information about the NANOG
mailing list