The MPLS WG Archive

[Paul Doolan <pdoolan@ennovatenetworks.com>]

>Paul> BGP ...   distributes reachability information  in the IDR  system. We
>Paul> may agree or disagree on how well it scales in that application.  That
>Paul> has little bearing on whether  that property applies in this different
>Paul> application.
>
Eric>Whether  something  scales in  an  extremely  large  environment has
little
Eric>bearing on whether it scales in  a smaller environment?  I can't imagine
why
Eric>anyone would make such a claim.

  Neither can I. And _again_ if you read a little more carefully you'll see that
it is
  not what I wrote.

  But since you mention it I'm a little puzzled by you choosing to define (the
NBVPN)
  as a 'smaller' environment. Lets assume the the (internet) core routers are
currently
  handling ~60,000 routes. And there are (?) thousands of AS's. That's big. No
doubt about
  it. A lot of routing information.

  In contrast  your own colleagues refer to the  "virtually
  unlimited scalability" of your VPN approach while your marketing  has talked
about
  "hundreds of thousands of VPNs". Allowing and order of magnitude  for
marketing
   hyperbole (and for the fact I can only find a reference to tens of thousands
  to hand :-) ) that leaves say 6 routes per VPN. Even with the paucity of
routes that I
  postulate per VPN this hardly seems to qualify as a 'smaller' environment.

  And we haven't begun to think about the poor folks who are being told that
they can
  use BGP to support VPNs while it is also performing the role for which it was
designed.
  And they are being told this......or at least not being clearly told the
opposite.

  pd




Eric Rosen wrote:

> I  don't really  want  to  keep it  up  either, but  as  long as  misleading
> statements are being made I feel I have to reply.
>
> Paul> BGP ...   distributes reachability information  in the IDR  system. We
> Paul> may agree or disagree on how well it scales in that application.  That
> Paul> has little bearing on whether  that property applies in this different
> Paul> application.
>
> Whether  something  scales in  an  extremely  large  environment has  little
> bearing on whether it scales in  a smaller environment?  I can't imagine why
> anyone would make such a claim.
>
> Paul> The scalability  of BGP is  only one of  the issues with  the protocol
> Paul> which call its use in this application into question.
>
> Now that's an example of FUD!!