6bone Architectural Changes? [RE: pTLA request for RMNET - review closes 23 April 2002]

[Joe St Sauver]

Hi,

>> The 6bone is _not_ real world. There is no real world, as of today. 

Sure there is. In fact, I would assert that IPv6 is very real real today,
and v6 is poised for accelerated availability (at least in the research 
and education network community) much in the way that IP multicast's 
diffusion had an inflection point a number of years back. (When native 
IP multicast was available and better than dvmrp connectivity, people 
moved... when native IPv6 is available and better than the 6bone, again, 
people will move)

If you look at http://monon.uits.iupui.edu/abilene/ipv6/ you'll see 
multiple IPv6 MRTG graphs running at >T1 speeds... as far as I'm concerned, 
traffic at that level's a sign that IPv6 is becoming quite real. 

>>When
>> we get native links, it will be time to re-assess the situation. (Bob, I
>> can provide you with a repeater in Sacramento when you lay out that dark
>> fiber between Berkeley and Truckee ;-)
>
>Native links are a sufficient but not required condition.  It would be
>sufficient that tunnels are only done locally "within a country or a
>region", or *at least* so that people with long tunnels don't use them for
>transit without thinking REALLY HARD about it.

Depending on what your tunnelled connectivity looks like and what your
native connectivity looks like, the tunnelled connectivity may (in some
cases) very well end up having lower latency then the native connectivity
-- it is all a function of the number and distribution of the native v6 
connectivity's peerings vis-a-vis the 6bone's interconnectivity 
(and the 6bone's aggregate interconnectivity is actually pretty good right 
now, I think, even if it *is* tunnelled). 

Consider Internet2's Abilene IPv6 service, for example. Because Abilene
currently relies exclusively on the 6tap for v6 peering with networks other 
than Abilene, traffic from Oregon to non-Abilene IPv6 destinations perforce 
HAS to travel to Chicago before it can go wherever its ultimate destination
may be. 

If I'm going from Oregon to IPv6 destinations in Europe, Chicago's "on the 
way," no harm, no foul, but if I'm going from Oregon to IPv6 destinations in 
Japan or Korea, say, I'd sure rather NOT have to go eastward just so I can 
turn around and go westward...

Likewise, if I lived on the East Coast instead of the West Coast, Chicago 
would be "on the way" for me going to Asian destinations, but an unnecessary 
detour if I was headed for Europe. If traffic has to make a big U turn to get 
delivered due to exchange point geographical distribution, I'd assert that 
there's a problem...

SO, from my point of view, if folks want to accelerate the migration of 
traffic from tunnelled connectivity to native connectivity, the key step is 
increasing participation at additional geographically distributed IPv6 
exchange points. http://www.6tap.net/ipv6-exchanges.html mentions a couple; 
see also the list at http://www.v6nap.net/

Alternatively, it would be terrific if Abilene and/or Canarie would offer an 
IPv6 International Transit Network service much as they currently offers an 
IPv4 ITN service for its non-US peer networks (for background on this, see:
http://www.ucaid.edu/abilene/html/itnservice.html ).

Regards,

Joe St Sauver (joe@oregon.uoregon.edu)
University of Oregon Computing Center