Re: [homenet] regarding recursive DHCPv6-PD

[Dave Taht <[e-mail]>]

On Thu, Nov 8, 2012 at 3:30 PM, Brian E Carpenter
<[e-mail]> wrote:
> Dave,
>
> On 08/11/2012 12:57, Dave Taht wrote:
>> On Thu, Nov 8, 2012 at 1:32 PM, Mikael Abrahamsson <[e-mail]> wrote:
>>> On Thu, 8 Nov 2012, Ted Lemon wrote:
>>>
>>>> On Nov 8, 2012, at 6:41 AM, Brian E Carpenter
>>>> <[e-mail]> wrote:
>>>>> Fine, but when such an end customer buys a second router and plugs it in,
>>>>> will she get an error message that says "Please find a new ISP"?
>>>>
>>>> In this case I think our only option is to fall back to bridging.
>>>
>>> Yes, doing protocol based brinding (L2 bridge 0x86dd packets) is the only
>>> way to go as far as I can tell.
>>
>> Um, er, ah, no, for many years now there has been the AHCP + babeld,
>> which routes /128s out of a /64 across any sort of wired/wireless
>> network over as many hops as needed.
>>
>> I've been doing my best to ignore this discussion, and work on
>> improving that code.
>
> Code is good, but where is the spec that we can refer to? As far as
> I can tell there's only an expired draft:
> http://tools.ietf.org/id/draft-chroboczek-ahcp-00.txt
> and RFC 6126, which is Experimental.

In both cases the current open source code reflects the drafts, and
has been in production use for several years.

Since the initial babel rfc, two interoperable implementations of the
babel protocol have arisen in quagga-RE and quagga mainline, in
addition to the standalone babeld daemon.

Additionally, cryptographic authentication of selected routes is now
available both as running code in quagga AND as ietf draft, updating
but not superceding rfc 6126, draft was posted in august:

http://tools.ietf.org/html/draft-ovsienko-babel-hmac-authentication-00

This makes it possible to withdraw certain networks from routing and
reduces the need for complex firewalling setups for ipv4 (natted or
non-natted) and ipv6.

Work is progressing on extending AHCP to do prefix distribution with a
similar algorithm to what is in the ospf version, but as noted,
shipping routed /128s around is simpler, and the default mechanism.


>
>>
>> You guys can beat your brains out over dhcp's approaches all you want,
>> and expect your ISPs in a cloudy future to deliver something bigger
>> than a /64 ...
>
> Some ISPs will do that much sooner.
>
>      Brian
>
>>
>> and bridge low speed wifi over high speed ethernet all you want -
>>
>> - but me, I kind of like being able to move transparently from AP to
>> AP and from wired to wireless and back again, and to not need anything
>> bigger than a /64 to do it.
>>
>> Given the trends towards excessive dynamicism throughout the IPv6
>> deployment, and the naming issues, and various ways proposed to
>> monetize/make scarce /sub 64 allocations by the ISPS. I think the
>> market will pick ipv6 nat and something like AHCP.
>>
>>
>>
>>



-- 
Dave Täht

Fixing bufferbloat with cerowrt: http://www.teklibre.com/cerowrt/subscribe.html