Re: [v6ops] Implementation Status of PREF64

[Ted Lemon <[e-mail]>]

I could successfully number all the devices in the network each with a /64
if I had a /59. This doesn't count IoT devices on IoT networks, I will
admit, but even then a /58 would do. What would not work would be to number
every device out of a /64.

That said, we're comparing apples to oranges here. On a home network, the
operator is /never/ going to configure DHCPv6. Sure, there are people here
who would, but they don't have a home network—they have a managed network
that is in their home. So yes, they would need a /56 or maybe even a /48
(but probably not).

On my home network, I just use SLAAC, and there's no problem. My Thread
network also does SLAAC (or whatever the Thread mesh does—I don't really
know, but definitely not DHCPv6).

So what other networks besides home networks are limited to a /56?

On Thu, Oct 14, 2021 at 3:07 PM Ole Troan <[e-mail]> wrote:

>
>
> On 14 Oct 2021, at 19:11, Ted Lemon <[e-mail]> wrote:
>
> 
> On Thu, Oct 14, 2021 at 2:56 AM Pascal Thubert (pthubert) <pthubert=
> [e-mail]> wrote:
>
>> A prefix to the host gives Lorenzo the addresses he needs for his
>> devices.  It gives your customers the single state per logical node that
>> they want. It allows to separate what netops manage (down to /64 and direct
>> assignment within) and devops (whatever they do with the longer prefix they
>> get for their node). It does not impose any size for what’s assigned, could
>> be a different thing for each host. It means routing inside the subnet
>> which removes the dreadful broadcast domain.
>>
>> I see an opportunity for consensus. Can we work that out together and
>> bring a real IPv6 value?
>>
>
> If your proposal is that we use a /64 per host as a way to meet these
> needs, I agree. This solves everybody's actual problems. There is the issue
> that some people have expressed a preference for prefixes wider than 64
> bits, but this is a preference—there's no technical reason to do this. It's
> not wrong to have preferences, but it would be nice if we could somehow
> finally put this discussion to bed.
>
>
> Limiting networks that are assigned a /56 to only 256 logical nodes is a
> non starter.
>
> O.
>