[5008] in North American Network Operators' Group

home help back first fref pref prev next nref lref last post

Re: Peering versus Transit

daemon@ATHENA.MIT.EDU (Alexis Rosen)
Fri Oct 4 03:59:23 1996

From: Alexis Rosen <alexis@panix.com>
To: alan@mindvision.com
Date: Fri, 4 Oct 1996 03:56:56 -0400 (EDT)
Cc: dorian@cic.net, barney@databus.com, nanog@merit.edu
In-Reply-To: <199610021600.LAA21703@anka.mindvision.com> from "Alan Hannan" at Oct 2, 96 11:00:58 am

Alan Hannan writes:
> [Alexis writes:]
> > It's a really bad decision. It saves the cost of hiring a real engineer, but
> > who wants to see a repeat of MAE-East? IXPs need a real traffic cop, at the
> > very least, to wreak havoc on people who play nasty link-layer games. (Yes,
> > it's conceivable that everyone on the IXP could guard themselves, but this
> > is highly inefficient both in dollars and hours spent.)
> 
>   It would seem to me that you've two rather positive choices ->
>   Elect the Routing Arbiter (Hi Bill :-) to police the XPs, or
>   through capitalism force the XP operators to implement such a
>   service.
> 
>   The former would be difficult as the're A/ overworked, and B/
>   officially powerless at the XPs (unless the XPs annoint them, which
>   is highly suspect).  The latter would be difficult in light of a
>   Robert Heinlein quote:
> 
>       " If you give the people the ability to vote themselves bread
>         and circuses, they will.  "

(Heh. I remember that quote.)

You've pretty much repeated my point...

>   Were the market to change (it might) we could have this.  Or, one
>   could create another market.  Several smallerish XPs (StLouiX)
>   comes to mind, have high quality peering standards built into
>   them.  I believe the CIX has done a fairly good job at this in the
>   past.
> 
>   But, it's my opinion that the only way to get MFS/PB/SL/AADS to
>   listen is with the pocketbook.

Thus my decision.

I think I'm going to tackle this one myself soon. I've given it a *lot* of
thought.

/a

home help back first fref pref prev next nref lref last post