> I have my doubts, based on a ~decade of observation. I don't think ARIN
> is deliberately evil, but I think there are some bits that'd be hard to
> fix.
I believe that anything at ARIN which the community at large and the membership
can come to consensus is broken will be relatively easy to fix.
Perhaps the true issue is that what you see as broken is perceived as "working
as intended" by much of the community and membership?
That's a great point. Would you agree, then, that much of the community
and membership implicitly sees little value in IPv6?
You can claim that's a bit of a stretch, but quite frankly, the RIR
policies, the sketchy support by providers, the lack of v6 support in
much common gear, and so many other things seem to be all conspiring
against v6 adoption. I need only point to v6 adoption rates to support
that statement.
This is an impediment that I've been idly pondering for some years
now, which is why I rattle cages to encourage discussion whenever I
see a promising opportunity.
Put differently, you work in this arena too... you've presumably
talked to stakeholders. Can you list some of the reasons people have
provided for not adopting v6, and are any of them related to the v6
policies regarding address space?
> Perhaps the true issue is that what you see as broken is perceived as "working
> as intended" by much of the community and membership?
That's a great point. Would you agree, then, that much of the community
and membership implicitly sees little value in IPv6?
Is that orthogonal to Owen's statement?
You can claim that's a bit of a stretch, but quite frankly, the RIR
policies, the sketchy support by providers, the lack of v6 support in
much common gear, and so many other things seem to be all conspiring
against v6 adoption. I need only point to v6 adoption rates to support
that statement.
In my experience ARIN/RIR policies have not been a noticeable barrier to
IPv6 adoption.
Lack of IA/security gear tops the list for my clients, with WAN Acceleration
a runner-up.
/TJ
> I have my doubts, based on a ~decade of observation. I don't think ARIN
> is deliberately evil, but I think there are some bits that'd be hard to
> fix.
I believe that anything at ARIN which the community at large and the
membership
can come to consensus is broken will be relatively easy to fix.
Perhaps the true issue is that what you see as broken is perceived as
"working
as intended" by much of the community and membership?
That's a great point. Would you agree, then, that much of the community
and membership implicitly sees little value in IPv6?
You can claim that's a bit of a stretch, but quite frankly, the RIR
policies, the sketchy support by providers, the lack of v6 support in
much common gear, and so many other things seem to be all conspiring
against v6 adoption. I need only point to v6 adoption rates to support
that statement.
This is an impediment that I've been idly pondering for some years
now, which is why I rattle cages to encourage discussion whenever I
see a promising opportunity.
Put differently, you work in this arena too... you've presumably
talked to stakeholders. Can you list some of the reasons people have
provided for not adopting v6, and are any of them related to the v6
policies regarding address space?
Perhaps the true issue is that what you see as broken is perceived as "working
as intended" by much of the community and membership?
That's a great point. Would you agree, then, that much of the community
and membership implicitly sees little value in IPv6?
I really don't know how much or how little value is seen in IPv6 by "much" of
the community. I see tremendous value in IPv6. I also see a number of
flaws in IPv6 (failure to include a scalable routing paradigm, for example).
Nonetheless, IPv4 is unsustainable going forward (NAT is bad enough,
LSN is even worse).
I do believe that IPv6 is being deployed and that deployment is accelerating.
I'm actually in a pretty good position to see that happen since I have access
to flow statistics for a good portion of the IPv6 internet.
The IPv6 internet today is already carrying more traffic than the IPv4
internet carried 10 years ago.
Many others see value in IPv6. Comcast and Verizon have both announced
residential customer IPv6 trials. Google, You Tube and Netflix are all
available as production services on IPv6. Yahoo has publicly announced
plans to have production services on IPv6 in the near future although they
have not yet announced specific dates.
I leave it up to you to consider whether that constitutes "much" of the
community or not.
Is that orthogonal to Owen's statement?
I don't see how the term orthogonal would apply here.
You can claim that's a bit of a stretch, but quite frankly, the RIR
policies, the sketchy support by providers, the lack of v6 support in
much common gear, and so many other things seem to be all conspiring
against v6 adoption. I need only point to v6 adoption rates to support
that statement.
IPv6 has had a slow start but it's certainly picking up.
IPv6 started approximately 20 years behind IPv4. It's already caught
up with IPv4 traffic levels of 10 years ago. Deployment is accelerating
and IPv4 will hit a sustainability wall in the near future.
Put differently, you work in this arena too... you've presumably
talked to stakeholders. Can you list some of the reasons people have
provided for not adopting v6, and are any of them related to the v6
policies regarding address space?
Reasons:
+ Fear
People simply fear deploying new technology to their environment.
+ Uncertainty
The future is uncertain. Many people fail to realize that IPv4's future
is even more uncertain than that of IPv6.
+ Doubt
You are not the only one expressing doubt in IPv6. The reality,
however, is that I think that LSN and a multi-layer NAT internet
are even more worthy of doubt than IPv6.
+ Inertia
Many people are approaching this like driving at night with the
headlights off. They refuse to alter course until they can see
the wall. There is a wall coming in two years whether you can
see it or not. If you have not begun to deploy IPv6 (changed
course), then there will soon come a point where the accident
has already occurred, even though you cannot yet see the
wall and have not yet made physical contact with it.
A classic example of this phenomenon would be a certain
large unsinkable ship where the captain chose to try and
make better time to New York rather than use a lower speed
to have time to avoid ice bergs. The ship never arrived in
New York and its name became an adjective to describe
large disasters.
+Bonus Fear: Because IPv6 deployments are small and vendors are still
ironing out software, there's concern that deploying it in a production
network could cause issues. (Whether or not this fear is legitimate
with vendor x, y, or z isn't the issue. The fear exists.)
+Bonus Uncertainty: There is a lack of consensus on how IPv6 is to be
deployed. For example, look at the ongoing debates on point to point
network sizes and the /64 network boundary in general. There's also no
tangible benefit to deploying IPv6 right now, and the tangible danger
that your v6 deployment will just have to be redone because there's some
flaw in the current v6 protocol or best practices that will be uncovered.
+Bonus Doubt: Because we've been told that "IPv4 will be dead in 2
years" for the last 20 years, and that "IPv6 will be deployed and a way
of life in 2 years" for the past 10, nobody really believes it anymore.
There's been an ongoing chant of "wolf" for so long, many people won't
believe it until things are much, much worse.
+Bonus Uncertainty: There is a lack of consensus on how IPv6 is to be
deployed. For example, look at the ongoing debates on point to point
network sizes and the /64 network boundary in general. There's also no
tangible benefit to deploying IPv6 right now, and the tangible danger
that your v6 deployment will just have to be redone because there's some
flaw in the current v6 protocol or best practices that will be uncovered.
This lack of consensus seems to most be associated with people who
haven't deployed. those of us who have in some cases a decade ago, don't
wonder very much...
You can deploy point-to-points as /112s or /64s. if you do anything that
isn't aligned on a byte boundary the brains will leak out of the ears of
your engineers. If you don't believe me go ahead and try it. any subnet
that has more than 2 devices on it is a /64 do anything else and you'll
shoot yourself or someone else in the foot and probably sooner rather
than later.
+Bonus Doubt: Because we've been told that "IPv4 will be dead in 2
years" for the last 20 years, and that "IPv6 will be deployed and a way
of life in 2 years" for the past 10, nobody really believes it anymore.
There's been an ongoing chant of "wolf" for so long, many people won't
believe it until things are much, much worse.
I bet you're really good at predicting the stock market as well. you can
be right and still go bankrupt. It is posisble to mistake postive but
nearly random outcomes for skill or insight.
I don't have to be right about needing an ipv6 deployment plan or even
believe that ipv6 is deployable in it's present form (I happen to
believe that, buts it's beside the point), because I need a business
continuity plan for what happens around ipv4 exhaustion, I may have more
than one, but I have a fiduciary duty to my company to not fly this
particular plane into avoidable terrain.
Put differently, you work in this arena too... you've presumably
talked to stakeholders. Can you list some of the reasons people have
provided for not adopting v6, and are any of them related to the v6
policies regarding address space?
Reasons:
(many excellent reasons removed)
Let me just add on:
(more excellent reasons removed)
I'm surprised no one has yet mentioned cost. Even if all the network equipment (hardware and software) in a given network were already v6 compatible, there's a substantial cost to train, test, document, deploy, support. Most companies will put this cost off for as long as possible, unless there are clear cost savings to be had by deploying sooner. Add in the problems getting vendors to produce v6 compatible networking equipment. Add in the cost to upgrade legacy systems to v6 compatible equipment (when available).
Most companies are trying to determine the optimum time to upgrade, and at this point they believe that this time is still in the future, not now.
Some of them will be up against a Y2K type of deadline when the v4 space runs out, scrambling to move to v6 when they need more IPs and can't get anymore usable v4 addresses.