Re: [v6ops] Request for your two cents: Pros and Cons of IPv6 Transition Technologies for IPv4aaS (464XLAT, DS-Lite, lw4o6, MAP-E, MAP-T)

[Lee Howard <[e-mail]>]

As both you and Fred said. . .


On 09/18/2018 08:50 PM, Mark Andrews wrote:
>
>> On 19 Sep 2018, at 1:23 am, Lee Howard <[e-mail]> wrote:
>>
>>
>>
>> On 09/17/2018 08:33 PM, Mark Andrews wrote:
>>> Requires using ISP’s recursive servers
>>>
>>> Yes, No, No, No, No
>>>
>>> We don’t yet have a NAT64 prefix discovery mechanism for 464XLAT that
>>> doesn’t use the DNS.  This is not the same as requiring DNS64.  The
>>> mapping can be in a locally served IPV4ONLY.ARPA zone (note this has
>>> all the same DNSSEC issues as DNS64 because IANA’s IPV4ONLY.ARPA zone
>>> is signed and securely delegated).
>>>
>> As I think about it, Google has public DNS64 servers that return a synthesized address using the WKP 64:ff9b::/96. So you might not be required to use the ISP's DNS (if the ISP uses the WKP).
>> Or, as you say, other DNS shenanigans.
> Which only works if your ISP is using the WKP which requires apiary knowledge (i.e. test your ISP’s nameservers or look it up on a web page).

Absolutely true.
Except I don't know what birds have to do with it.

I'm only pointing out that there is a way out of DNS capture by the ISP, 
so it doesn't *require* ISP's recursive resolvers, except in every 
instance known in production.

Lee