Re: draft-bourbaki-6man-classless-ipv6-00

[David Farmer <[e-mail]>]

On Sat, Jun 3, 2017 at 12:32 PM, Roger Jørgensen <[e-mail]> wrote:
>
>
> Things like this ONLY hurt IPv6, yet again IETF can't agree on IPv6 so why
> should we bother using it? But I guess the none technical side are a lost
> case for most IETF'ers. Not to mention that we remove one of the thing
> so many I've spoken to over the years think is great - the standard LAN
> size. No more discussion on that.
>

Personally, I'm fine with a standard LAN subnet size, but that's not what
RFC4291 says, it says /64 is magic and all subnets are /64 regardless.
However, not every network in the world is a LAN. Just because we use
ethernet framing everywhere doesn't make every network a LAN either. We
have added /127 with RFC6164 for point-to-point, but if you use a /30 for
IPv4 point-to-points, then a /126 makes more sense for your IPv6
point-to-points.

Anyway, changing to "/64 is the RECOMMENDED subnet size" is still provides
a standardized LAN size. It just recognizes that not everything is a LAN.

More generally, I understand the desire to break with the old conventions
from IPv4, and eventually we have to. However, right now we are in a
transition state where we have to support both IPv4 and IPv6, and to much
divergence in the operational models between IPv4 and IPv6 just creates
issues.  Yes this costs money and the bean counters don't like that, but
even more importantly it reduces reliability and lowers the quality of the
user experience.  This is what you call a lose-lose situation.

And talk about hurting IPv6, if deploying it reduces reliability and and
lowers the quality of the user experience for a network, that's what will
hurt IPv6.

Finally, saying all LANs are /64 only works if providers had out prefixes
to customers that are larger than /64. In my experience that's not
happening, if you know how to fix that please speak up!  Otherwise, we
better figure out have to split a /64 across multiple local subnets,
because its very difficult for lay people to understand why 18
quintillion addresses isn't enough for their house, no matter how many
networks they have in their house.

Thanks.



-- 
===============================================
David Farmer               Email:[e-mail]
Networking & Telecommunication Services
Office of Information Technology
University of Minnesota
2218 University Ave SE        Phone: [phone] <(612)%20626-0815>
Minneapolis, MN 55414-3029   Cell: 612-812-9952 <(612)%20812-9952>
===============================================