Alternative Re: ipv4/25s and above Re: 202211201009.AYC

Sun Nov 27 23:49:37 UTC 2022

On Mon, Nov 21, 2022 at 4:05 PM David Conrad <drc at virtualized.org> wrote:
>
> Barry,
>
> On Nov 21, 2022, at 3:01 PM, bzs at theworld.com wrote:
>
> We've been trying to get people to adopt IPv6 widely for 30 years with very limited success
>
>
> According to https://www.google.com/intl/en/ipv6/statistics.html, it looks like we’ve gone from ~0% to ~40% in 12 years. https://stats.labs.apnic.net/ipv6 has it around 30%. Given an Internet population of about 5B, this can (simplistically and wrongly) argued to mean 1.5-2B people are using IPv6. For a transition to a technology that the vast majority of people who pay the bills will neither notice nor care about, and for which the business case typically needs projection way past the normal quarterly focus of shareholders, that seems pretty successful to me.
>
> But back to the latest proposal to rearrange deck chairs on the IPv4 Titanic, the fundamental and obvious flaw is the assertion of "commenting out one line code”. There isn’t “one line of code”. There are literally _billions_ of instances of “one line of code”, the vast majority of which need to be changed/deployed/tested with absolutely no business case to do so that isn’t better met with deploying IPv6+IPv4aaS. I believe this has been pointed out numerous times, but it falls on deaf ears, so the discussion gets a bit tedious.

I have been trying to steer clear of this debate this time around, but
since I'm the one that made that analogy to begin with...

There are now billions and billions of *non-instances* of this one
line of code, saving nanoseconds on every connection, since 2008 in
the case of 240/4 and 2018 in the case of 0/8 - and that savings
alone, I felt was worth it. No additional future use is required from
my perspective to have realized real economic value from these address
spaces.

It would be rather nice, if, over time, we pretty much agreed that
embedding an 1981 policy into future OS kernels and routers transport
mechanisms was silly.

Full stop. Can someone citing me about the non-wisdom of "delete 1
line of code from everything" try to explain why our OSes MUST
continue enforcing some distinction between 240/4 and 0/8 and the rest
of the known unicast internet?

...

To take the next step - towards some sort of allocation policy - is a
matter of years and years. The subject of current research is what
does trying to make it work, break? I regularly use 240 nowadays
myself where I am not sure where the rfc1918 space is... or on a vpn -
eating my dogfood - but I do think it would be a tragic waste if we
didn't make an effort to make them globally usable in the long run.

I also tend to be upset by the argument that "this must work
internet-wide, on everything, forever, and immediately", which of
course, doesn't apply to ipv6 either.

No, it just needs to work on islands with limited address space,
initially. Tunnels between forward thinking providers, perhaps.
Starlink could use it to address terminals if they wanted - they still
don't have ipv6 working worth a darn -

I've also said a lot, that "the prospect of a portion of the internet
completely immune to windows-born viruses and worms is really
pleasing..." and I get a lot of laughs from that, because it's true -
If you've been in the trenches, fighting those off for the last few
decades, knowing that *some* piece of your infrastructure couldn't be
subject to those sort of attacks from old or windows OSes is a relief.

Arguments for deploying ipv6 remain! There's no escaping ipv6, and I
spend a lot of time trying to convince ISPs nowadays to deploy that,
but *all* of the ones I'm presently working with still must provide
IPv4 space, and thus are deploying CGNAT more rapidly than ipv6. I
will keep trying to get ipv6 deployed at every chance I get! I'm very
happy to have finally got ipv6 trie support into libreqos.io a few
weeks ago - but the demand is all cgnat, and mpls and vlans and ipv4
tunnels - I'd love to find a customer to try out our new ipv6 support,
because despite trying for months, we don't have any, as yet.

Blatant plug: https://github.com/LibreQoE/LibreQoS/tree/main/v1.3#v13-ipv4--ipv6-beta

Anyway... some use of these new ipv4 address spaces is inevitable, and
I really do wish y'all cared more about nanoseconds,
here or there, or anywhere.

>
> Regards,
> -drc
>

-- 
This song goes out to all the folk that thought Stadia would work:
https://www.linkedin.com/posts/dtaht_the-mushroom-song-activity-6981366665607352320-FXtz
Dave Täht CEO, TekLibre, LLC