[6bone] separating IPv6 experimental from production traffic

[Gert Doering]

Hi,

On Thu, Aug 22, 2002 at 04:17:46PM +0300, Pekka Savola wrote:
> On Thu, 22 Aug 2002, Gert Doering wrote:
> > "Full transit to everybody" was a good idea to get a pretty tightly 
> > meshed IPv6 network into place, which is *good* because it means that
> > you have (typically) only few AS hops to traverse - and thus fewer 
> > networks in between that can mess up your routing.
> 
> And quality (or lack thereof) is equally indeterminate everywhere; as-path
> length tells *nothing* about optimal paths, ...

Of course.  But if you have *enough* peerings, you'll be able to reach
most networks in a maximum of 2 hops - and if you then apply some 
MED fiddling to mark "slow" tunnels, you can achieve pretty good results.

(Of course it's helpful if you tag your routes accordingly, so your peers
can filter accordingly).

This works amazingly well :-)

> Hierarchy is good.

Hierarchy isn't too bad *iff* there are major differences between 
participants.

As long as none of my upstream ISPs *offer* IPv6 connectivity, I don't 
see any reason why some of the ASes out there should be "further up"
or "further down" in the hierarchy than we are.

They are *peers*, more or less.

[..]
> The question is whether we want to evolve 6bone network (very difficult,
> but only if some transits would step up, *some* process might evolve..)  
> or let it rot on purpose.  I'm starting to think voting for b).

I see this as a market issue.  As soon as someone is actually willing to
spend money on *good* IPv6 connectivity, there is an incentive to avoid 
routing over trans-continental (or even international) tunnels, and things
will unravel on their own.

Of course I'm all for adding native peerings, dropping tunnels to
non-responsive peers (or to some that have a dendency to blackhole 
things), and so on.  But I don't expect a change to a hierarchically 
organized network of purely native links "really soon now".

Gert Doering
        -- NetMaster
-- 
Total number of prefixes smaller than registry allocations:  45583  (46871)

SpaceNet AG                 Mail: netmaster@Space.Net
Joseph-Dollinger-Bogen 14   Tel : +49-89-32356-0
80807 Muenchen              Fax : +49-89-32356-299