North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

Re: ANNOUNCEMENT: NANOG 9 Date Change (fwd)

  • From: Robert Laughlin
  • Date: Tue Nov 26 11:19:25 1996

I vote for Avi description of the topic.

Robert

----------------------------------------------------------------------------
DataXchange                sales:  800-863-1550            http://www.dx.net
       Network Operations Center:  703-903-7412 -or- 888-903-7412
----------------------------------------------------------------------------

On Tue, 26 Nov 1996, R. Eric Bennett wrote:

> > 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.         [email protected]
> >
> > 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 <[email protected]>       |   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
> 
> 
> 
- - - - - - - - - - - - - - - - -