6BONE shutdown

Iljitsch van Beijnum iljitsch at muada.com
Tue May 9 22:14:44 CEST 2006

On 9-mei-2006, at 21:38, Doug Barton wrote:

> Your point about folks preferring to do nothing is well taken. This  
> leads me
> to two conclusions. First, it is good to remove your 6bone routes  
> because
> they are likely not to work, or worse, only work "sometimes."  
> Second, those
> who feel that deprecation of the 6bone is a goal should do  
> everything in
> their power to dial up the pain for those still sending traffic  
> with those
> addresses on 7 June.

As a final note on this topic (and the other one as well, I think)  
let me observe that it's often not the source of the packets that's  
in direct control of the destination address. For instance, I  
currently have both production and 6bone addresses in the DNS for my  
personal site www.muada.com (removed 6bone addresses from my other  
sites some months ago). So if you point your IPv6-enabled web browser  
towards that domain you have a pretty good chance of hitting the  
6bone address. Now obviously if you have an IPv6 implementation that  
supports a policy table for address selection, such as FreeBSD 6 or  
Windows XP, you can set things up so that 6bone space is avoided, but  
I'm guessing most people won't do that. (ip6addrctl in FreeBSD, but  
enable it first in /etc/rc.conf and netsh interface ipv6 add|show  
prefixpolicy on XP with some caveats.)

In any event, I'll leave 6bone.muada.com in place for testing  
purposes until connectivity towards my upstream for the address in  
question stops. This name doesn't have an IPv4 or non-6bone IPv6  
address so if you can ping it or hit it with your browser you know  
packets can still flow from you to the 6bone, and if not, not.

> If IPv6 is ever going to be considered more than an interesting  
> idea, it has
> to start being treated as a production network.

Yes. But note that for various reasons, IPv6 will continue to be  
different from IPv4 in many minor but also in some major aspects.

More information about the ipv6-ops mailing list