[arin-ppml] Routing Research Group is about to decide its scalable routing recommendation

[Davis, Terry L]

Leo

PA assumes that we discover how to keep our developers and ops staffs from embedding low level network infrastructure into their applications and processes.  And that so far we have not managed to do; so from a CIO's point of view, PA is inherently hard to sell due the risk of system/application outages involved with changing ISPs.

Most businesses also require multi-homing to meet their "business continuity" requirements; so to a CIO of any sizable company/entity, PA without it is a non-starter.

Also I understand that the risks associated either of the above can become an issue in SOX compliance reporting.

Take care
Terry

> -----Original Message-----
> From: arin-ppml-bounces at arin.net [mailto:arin-ppml-bounces at arin.net] On
> Behalf Of Leo Bicknell
> Sent: Friday, December 18, 2009 11:51 AM
> To: arin-ppml at arin.net
> Subject: Re: [arin-ppml] Routing Research Group is about to decide its
> scalable routing recommendation
> 
> In a message written on Fri, Dec 18, 2009 at 10:34:58AM -0800, Michel Py
> wrote:
> > > Leo Bicknell wrote:
> > > What I am hearing is that the research community believes that BGP4
> > > cannot scale to the point where everyone has a provider independent
> > > prefix.  That if we want to have everyone have a provider independent
> > > prefix, we need a new technology to be in place first.
> >
> > That technology was supposed to be IPv6, PA-only. Unfortunately it was
> > launched before the quest for a working PA-only multihoming solution
> > ended up returning with nothing.
> 
> I agree that was the IETF's direction, but it is also a different
> solution.  As I see it, the problem space has three, high level,
> diverging paths:
> 
> A Multihoming happens entirely at the host, making PA-only possible.
> 
> B The routing system can scale to PI, so everyone has PI.
> 
> C Neither A or B is possible, so we attempt to decide who is worthy
>   of PI.
> 
> It seems to me we are in case C now, the IETF tried A several years ago
> and gave up, and the IETF is now trying B.  (roughly)
> 
> Which raises an interesting question, why hasn't SCTP taken off more?
> 
> http://en.wikipedia.org/wiki/Stream_Control_Transmission_Protocol
> 
> --
>        Leo Bicknell - bicknell at ufp.org - CCIE 3440
>         PGP keys at http://www.ufp.org/~bicknell/