How many Tunnels [Re: Re: pTLA request]

[Buclin, Bertrand]

> Is there any opinion on mentioning these problems in the "6bone
> Routing Practice" I-D, not as a recommendation but as a problem
> statement?
> 
Simon,

You're making a set of good points. The I-D currently recommends that pTLAs
have at least 3 peerings with other pTLAs to maintain a good level of
service. This was discussed on the mailing list back in April.

If some pTLAs want to maintain a larger set of peerings, this is fine by me
(I'm actually one of those maintaining a larger set, currently we have about
12 peerings with other pTLAs). There are benefits in doing this to:

- I do this because I have nodes both in Europe and in the US, and this way
I can provide a transit service and experiment with what it takes. UUNET-UK
is in the same situation. I agree though with your point if a pTLA has only
one site, in that case establishing peerings with entities all over the
world is counter productive.
- Abundance of tunnels also allows us to work on the issue raised by the
wider propagation of routing problems. This is particularly true if  you
operate a distributed pTLA.
- Ideally, having a 6bone topology matching the underlying IPv4 Internet
topology is a neat objective. We've been trying to do that for a while with
little success. Remember that the 6Bone connects entities doing research on
Ipv6, not yet operators.

CHeers, Bertrand